# Blog
Source: https://docs.automatstudio.com/blog/index

Product updates, releases, and news from Automat Studio

#### Thu, Sep 17, 2026

<img alt="Reference images converging into one unified graded frame" />

**Revision Is Directing**

A first version is a proposal. Nobody hangs the first lighting setup, keeps the first costume fitting, or prints the first take — not because first attempts are bad, but because the director's real work only starts once there's something to react to. The work is the note.

Anyone can accept or reject. Directing is the third option: adjust. This take, but slower. This costume, but worn like she's owned it for years. This room, but colder. Every great collaborator you'll ever hear about — the cinematographer, the editor, the production designer — is someone who turns precise notes into precise changes.

The skill is in the precision. "Make it better" is a wish; "let the pause land before she answers" is a direction. When you catch yourself giving a vague note, it's rarely a vocabulary problem — it's a sign you haven't decided what the moment is for. The note forces the decision. That's why revision isn't the boring part after creativity: it's where the actual directing happens.

In Automat Studio, your whole world takes notes now — characters, locations, and props revise from a sentence, and the new Edit workflow keeps everything you didn't mention. Say exactly what you want changed. Deciding that is the job. [https://www.automatstudio.com/](https://www.automatstudio.com/)

***

#### Mon, Sep 14, 2026

<img alt="The same character before and after one described change" />

**Change Anything, Keep Everything**

Every character, location, and prop in Automat Studio can be revised with a sentence.

Open the asset, switch to Edit, and describe the change: "give her a scar above the left eye." "Make the café rain-soaked and closed for the night." "Age the sword a hundred years." The image updates with the change you asked for — and everything you didn't mention stays exactly as it was.

This is the difference between generating and directing. Rerolling a character is a gamble: you might fix the jacket and lose the face. Giving a note is a decision: the person who comes back is still the person you cast. Combined with the other two ways of building an asset — describing from scratch, or generating from reference images — Edit closes the loop: create, look, revise, exactly the rhythm of real pre-production.

Your cast can take notes now. Give them one: [https://www.automatstudio.com/](https://www.automatstudio.com/)

***

#### Thu, Sep 10, 2026

<img alt="A camera dolly on curved track mid-move through a small film set" />

**The Edit Begins Before the Shoot**

Editors like to say the film is made three times — once when it's written, once when it's shot, once when it's cut. It's true, and it hides a better secret: the directors who suffer least in the cutting room are the ones who were already editing before anything was shot.

Editing-before-shooting isn't about locking yourself in. It's about asking the editor's questions early, when answers are cheap. Which scene follows which? What does the audience know at this exact moment, and when do they learn the next thing? Where does the breather go after the blow lands? On paper, restructuring costs seconds. On set it costs a day. In the edit it costs the scene you never shot.

Your scene list is a rough cut in disguise. Read it top to bottom the way a viewer would meet it. If the middle sags in outline, it will sag worse with pictures attached, because pictures are slower than prose. Fix the sag now, while the fix is free — then fix it again. A structure that survives fifty cheap revisions has earned its shape.

That's why restructuring in Automat Studio is drag-and-drop: the cost of asking "what if the reveal came first?" should be nothing at all. [https://www.automatstudio.com/](https://www.automatstudio.com/)

***

#### Mon, Sep 7, 2026

<img alt="A tidy film production office with scene cards, monitors and binders" />

**Your Studio, Organized**

A film is a production before it's a picture — and lately Automat Studio has quietly become a much better production office.

The Navigate menu puts every department one click away from anywhere: script, characters, locations, wardrobe, props, scenes, shots. Scenes drag and drop to reorder, so restructuring your story costs seconds instead of ceremony. Titles rename inline — click, type, done. And generations you're not using hide away without being deleted: your working set stays clean, your history stays intact.

None of these is a headline feature, and that's the point. Production management is the craft nobody applauds and every finished film depends on. The less friction between "I should restructure this" and having restructured it, the more drafts your story survives.

Run your studio: [https://www.automatstudio.com/](https://www.automatstudio.com/)

***

#### Thu, Sep 3, 2026

<img alt="A director's hand hovering at a glowing monitor in a dark studio" />

**Hold the Shot**

Most cuts happen because the editor got nervous, not because the shot was finished. A held shot builds pressure. The audience, given time, starts looking deeper into the frame — reading hands, doorways, the space behind the actor. They begin doing the scene's work with you. Cut too soon and you release that pressure for nothing.

The great long takes understand this. The opening of Touch of Evil winds a bomb through a border town and refuses to look away. Children of Men stages catastrophe in single breaths. The tension isn't in what happens — it's in how long you're made to wait for it, and in the growing suspicion that no one is coming to rescue you with a cut.

The rule is simple to say and hard to obey: cut when the meaning changes, not when the motion stops. If the shot is still asking a question, stay. If it has answered, leave before the audience does.

With single takes now running up to thirty seconds in Automat Studio, the long take is a choice you can actually make. Make it on purpose. [https://www.automatstudio.com/](https://www.automatstudio.com/)

***

#### Mon, Aug 31, 2026

<img alt="A steadicam operator following two actors through a golden-hour long take" />

**The 30-Second Take**

Thirty seconds is a different creative unit. Four seconds is a beat; thirty is a scene — a walk-and-talk, a slow push down a corridor, a conversation that's allowed to breathe.

In Automat Studio, choose Seedance 2.5 for your shot, set any duration from 4 to 30 seconds, and generate one continuous take. No stitching, no hidden cuts, and sound comes included at no extra cost. It works from a still frame or from a direct prompt, and Performance Capture takes run from 2 to 30 seconds too — so a long performance carries over whole.

The long take has always been filmmaking's trust exercise: the director trusting the staging, the audience trusting the frame. Now it's one prompt away.

Try it today: [https://www.automatstudio.com/](https://www.automatstudio.com/)

***

#### Thu, Aug 27, 2026

<img alt="The same street corner rendered in Film Noir and Painterly Neon" />

**A Look Is a Promise**

When a film commits to a look — a palette, a lens, a quality of light — it makes the audience a promise: this world has rules, and we know what they are. Every consistent frame renews that promise. Every random one breaks it.

That's why the films you remember feel inevitable. The Godfather's amber shadows. Amélie's saturated greens. Mad Max's chrome and dust. You could identify a single frame from across a room, because every department — production design, wardrobe, lighting, grade — was serving one visual idea.

The promise matters most when resources are small. A short film with a committed look reads as intentional; the same film with a drifting one reads as accidental, even if every individual shot is beautiful. Consistency is doing the work that budget can't.

This is exactly what aesthetics are for in Automat Studio: choose the world's rules once, and every shot obeys them. Then, when you finally do break the look — a single scene in color, one handheld shot in a locked-off film — the audience feels it, because you spent the whole film earning it.

Choose the look first. Keep it. Break it only on purpose. [https://www.automatstudio.com/](https://www.automatstudio.com/)

***

#### Mon, Aug 24, 2026

<img alt="All twelve free aesthetics in Automat Studio, one card each" />

**Twelve Looks, One Click**

We're excited to announce that free aesthetics in Automat Studio have doubled — from six to twelve.

An aesthetic is your film's look, applied everywhere at once: era, genre, film stock, palette. Choose it when you create a project — or change it in settings — and every shot you generate renders in the same world. It's the difference between a folder of clips and a film.

The six new looks: **Painterly Neon**, mixed-media comic-book animation with visible brushwork; **Watercolor Anime**, hand-painted backgrounds and cel-shaded characters; **Film Noir**, 1940s black-and-white with hard shadows and rain-slicked streets; **Storybook Symmetry**, dead-centre framing and candy-box pastels; **Stop-Motion**, the handcrafted clay-and-felt miniature look; and **Retro VHS**, 80s camcorder nostalgia complete with scanlines and chroma bleed.

All twelve are available on every plan, including free. The same drop refreshed five of the original six with new artwork and richer style descriptions.

Try them today: [https://www.automatstudio.com/](https://www.automatstudio.com/)

***

#### Thu, Aug 20, 2026

<img alt="Five stills showcasing the new July models" />

**Five New Models Land in Automat Studio**

Five new models landed in Automat Studio this month.

Kling O3 brings top-tier image-to-video motion. Gemini Omni Flash generates video with native audio and supports multi-turn editing. Wan 2.7 delivers fast, expressive video generation. Seedream V5 Pro is our new benchmark for cinematic stills. And Nano Banana 2 Lite gives you rapid image drafts before you commit credits.

Every model is rated in-app so you always know which are recommended for the job — and your whole roster works inside one pipeline: same characters, same locations, same continuity.

Explore the roster: [https://www.automatstudio.com/](https://www.automatstudio.com/)

***

#### Tue, Aug 18, 2026

<img alt="The same character in three different costumes" />

**Introducing Wardrobe**

Costume designers know something the rest of us learn late: clothes are backstory you can see.

Wardrobe is now a first-class part of Automat Studio. Design wardrobe items once — the rain coat, the borrowed suit, the armor — and @-mention them in any shot. Your character wears the same coat in scene 3 and scene 30, because the coat is an asset, not a description.

Costume continuity, solved the same way we solved it for characters, locations, and props.

Dress your cast: [https://www.automatstudio.com/](https://www.automatstudio.com/)

***

#### Thu, Aug 13, 2026

<img alt="A kitchen at dusk with a single chair at the table" />

**Let the Space Tell the Story**

Before a character says a word, the room has already spoken. A kitchen with one chair. An office where the blinds are always drawn. A child's bedroom kept exactly as it was left. Audiences read spaces instantly and involuntarily — production design is exposition that nobody has to say out loud.

When you design a location, don't ask what it looks like. Ask what happened here, and what the room is still holding onto.

Design spaces that speak in Location Studio: [https://www.automatstudio.com/](https://www.automatstudio.com/)

***

#### Tue, Aug 11, 2026

<img alt="The same cafe seen from four camera angles" />

**Introducing Location Views**

A film location isn't one image. It's a place you can shoot from anywhere.

Location Views brings that to Automat Studio: generate multiple angles of the same location — the wide from the door, the corner booth, the view toward the window — all geometrically consistent, all recognizably the same room.

Block a whole scene inside one coherent space. Reverse shots that actually reverse. Coverage that cuts together because the geography holds.

Your locations are now places, not pictures. Try it today: [https://www.automatstudio.com/](https://www.automatstudio.com/)

***

#### Thu, Aug 6, 2026

<img alt="A wall of storyboard frames under warm light" />

**Think in Sequences, Not Shots**

A beautiful shot that doesn't cut is a postcard. Films are built from sequences — shots that need each other. The wide means something because the close-up answers it. The empty hallway matters because someone is about to enter it.

A shot rarely works alone. Meaning comes from what precedes and follows it: a glance lands harder after a closed door, a wide feels lonely after a close-up. When you plan a scene, think about how each shot hands off to the next — movement, eyelines, and energy should carry across the cut.

Before generating a single frame, sketch the sequence: what does each shot ask, and which shot answers it? If a shot doesn't ask or answer anything, it doesn't belong.

Beautiful is easy now. Necessary is still the craft.

Build your sequences in Automat Studio: [https://www.automatstudio.com/](https://www.automatstudio.com/)

***

#### Tue, Aug 4, 2026

<img alt="The same street at golden day and blue dusk" />

**Runway Aleph 2 Has Arrived in Automat Studio**

Runway Aleph 2 is now live in Automat Studio.

Aleph 2 is a video-to-video model: give it existing footage and direct the change. Turn day into dusk. Swap the season. Restyle the entire frame while the motion and blocking stay exactly as you shot them.

For filmmakers, this is a new kind of coverage — the take you got becomes the take you needed. It's another way Automat Studio lets you make creative decisions late, without going back to set.

Try it today: [https://www.automatstudio.com/](https://www.automatstudio.com/)

***

#### Thu, July 30, 2026

<img alt="A filmmaker's desk with storyboards and a glowing monitor" />

**Keep Human Creativity in Filmmaking**

Every tool we add to Automat Studio is measured against one question: does the filmmaker stay in charge?

AI can render a frame, but it can't want anything. It doesn't know why the scene matters, what the silence is hiding, or which imperfection to keep. Those judgments are the film. They belong to you.

That's why Automat Studio is built like a studio, not a slot machine. Characters you cast. Locations you scout. Shots you compose, in sequences you structure. And now performances you act yourself, carried onto any character you've created.

The models will keep getting better. Our promise stays the same: they render, you direct.

The story is always yours. [https://www.automatstudio.com/](https://www.automatstudio.com/)

***

#### Mon, July 27, 2026

<img alt="A performer and the fantasy character delivering the same performance" />

**Introducing Performance Capture**

We're excited to announce Performance Capture, now live in Automat Studio.

Record a performance — on your phone, in a rehearsal room, anywhere. Upload it to Shot Studio, and your character delivers it. The gestures, the timing, the way a pause lands: it's all carried onto any character in your film.

Acting has always been the hardest thing to fake with AI. So we stopped trying to fake it. The performance comes from a person. The character comes from your studio. Automat Studio brings them together.

Try it today: [https://www.automatstudio.com/](https://www.automatstudio.com/)

***

#### Thu, Feb 19, 2026

<img alt="A triptych of a young woman at a cafe" />

**A Pipeline That Ties It All Together**

Automat Studio integrates 100+ AI models that generate images, video, and audio.

Used separately, these tools are powerful but fragmented. You have to manage files, download media, re-upload assets, track versions, and build your own workflow across different providers.

Automat Studio handles that workflow for you. Your images, videos, and audio all live inside a single structured studio. Characters stay connected to their visuals. Locations remain consistent. Shots build into scenes. Media is organized automatically as your project evolves.

Instead of stitching tools together yourself, Automat Studio connects the models into one cohesive filmmaking pipeline.

Try it now: [https://www.automatstudio.com/](https://www.automatstudio.com/)

***

#### Wed, Feb 11, 2026

<img alt="A grid of characters" />

**Introducing Reference Images**

You can now upload reference images in Character Studio, Location Studio, and Prop Studio. This has been one of the most requested features, and it gives you more precise control over your characters, locations, and props, bringing the results closer to your vision.

Instead of describing everything from scratch, you can now show the system what you mean. A face, a room, an object. The image becomes part of your creative direction.

Try it now: [https://www.automatstudio.com/](https://www.automatstudio.com/)

***

#### Wed, Feb 4, 2025

<iframe title="Prop Studio Tutorial" />

**Every Object Tells a Story**

In film, props are never just decoration. They reveal character, suggest history, and quietly carry meaning from one scene to the next.

Introducing Prop Studio, a new way to design and reuse props with intention, keeping objects consistent across shots and scenes as your project evolves. It is another step toward bringing real filmmaking structure into your creative workflow.

Try it today: [https://www.automatstudio.com/](https://www.automatstudio.com/)

***

#### Thu, Jan 29, 2026

<img alt="A young filmmaker sits at a computer creating a film" />

**Automat Studio is designed to take your time.**

Automat Studio is designed to take your time. This is not just a prompt box. It is a filmmaking studio built around a structured production pipeline.

You craft ideas carefully and precisely, the way films are traditionally made. Characters, locations, dialogue, and performance are developed step by step, refined over time, and kept consistent as your project evolves.

Automat Studio follows the natural flow of filmmaking, from concept to scene to shot. It gives you space to iterate, make deliberate choices, and shape a film that reflects your vision.

If you value control, continuity, and a real creative process, Automat Studio was built for you.

Explore Automat Studio: [https://www.automatstudio.com/](https://www.automatstudio.com/)

***

#### Tue, Jan 20, 2026

<iframe title="Scenes and Shots Tutorial" />

**Define the Look of Your Film with Aesthetics**

Visual consistency is one of the hardest things to get right in filmmaking, especially when you’re working fast.

In Automat Studio, Aesthetics let you define the look and feel of your entire project. Choose from built-in styles like European arthouse or Dark Academia, or create your own custom aesthetic using reference images. Once applied, every new image you generate follows that visual language automatically.

We just released a short tutorial showing exactly how aesthetics work and how to create your own in minutes.

Watch the tutorial and give your projects a coherent visual identity.

***

#### Tue, Jan 13, 2026

<img alt="A man sits at a computer creating a film" />

**Make Your Film Before You Make Your Film**

Automat Studio lets you visualize a film end to end before the camera ever rolls.

It’s a space to explore ideas, see characters and locations take shape, test shots, and iterate freely. By the time you step on set, the creative questions have already been asked and answered, leaving you to focus on the moment itself.

Try it now: [https://www.automatstudio.com/](https://www.automatstudio.com/)

***

#### Tue, Jan 6, 2026

<iframe title="Scenes and Shots Tutorial" />

**The Easiest Way to Create Shots and Scenes**

Scenes and shots are the building blocks of any film. Automat Studio is structured around those same ideas, bringing traditional filmmaking concepts into an AI-driven workflow.

This short tutorial shows how to create scenes and build shots within them, keeping your work grounded in familiar cinematic structure while you experiment visually.

It’s a quick walkthrough, designed to help you move from ideas to images without friction.

***

#### Tue, Dec 30, 2025

<img alt="A man sits at a computer creating a film" />

**Who is Automat Studio for?**

Automat Studio is for anyone who wants to bring a film idea to life. You can use it as a pre-visualization tool to explore scenes, test ideas, and iterate before cameras ever roll. Or you can use it to create the film you have always imagined but never had the budget, crew, or access to make. It is a way for ideas that might have stayed in your head, or in a notebook, to finally take shape.

Automat Studio is for anyone with a creative vision.

Make Your Film Now: [Automat Studio](https://automatstudio.com).

***

#### Tue, Dec 16, 2025

<iframe title="Location Studio Tutorial" />

A foggy mountain cabin?
A dimly lit jazz club?
A quiet Tokyo alley at midnight?

Introducing Location Studio. The place where you design the environments that shape your film’s world. From intimate interiors to expansive landscapes, Location Studio lets you define the mood, atmosphere, and visual language of every setting before a single shot is created.

If you can imagine it, you can start building it.

Try it today: [https://www.automatstudio.com/](https://www.automatstudio.com/)

***

#### Tue, Dec 9, 2025

<img alt="A couple dancing" />

**Video Generation Is Now Available to Free Users**

Free tier users can now create videos directly from the Video tab in Shot Studio. No subscription is required. Your free allowance is enough to try two or three videos and get a feel for the workflow, along with additional images for your characters and locations.

We have also improved prompt adherence in video generation so your shots follow your creative direction more closely.

To help you explore even further, we have doubled the free allowance for all new users and anyone who signed up recently.

Try video generation now: [Automat Studio](https://automatstudio.com).

***

#### Tue, Dec 2, 2025

<img alt="Two filmmakers watching a video monitor" />

**Automat Studio is structured around the traditional filmmaking process.**

We’ve designed Automat Studio to reflect the roles of an actual film production. Character Studio is your casting department. Location Studio handles locations and set decoration. Aesthetics covers cinematography. Shot Creator lets the director compose the frame. More departments are coming soon.

Filmmaking has always been a collaborative craft. With Automat Studio, we’re preserving that structure — while making it easier to experiment, iterate, and bring your vision to life.

If you know how films are made, Automat Studio will feel familiar. It’s built on the same creative foundations.

Explore it today at [Automat Studio](https://automatstudio.com).

***

#### Mon, Dec 1, 2025

**Kling Video O1 Now Available in Automat Studio**

<img alt="A young woman sitting in a cafe" />

We’re excited to announce that Kling Video O1 is now live in Automat Studio.

Kling’s new model delivers incredibly sharp motion, detailed textures, and some of the most stable video generation we’ve seen yet — and it’s available exclusively through the fal.ai API. Now you can use it directly inside Shot Studio to create smoother, more cinematic camera movement and richer visual detail.

If you’ve been looking for a model that handles action, dynamic shots, and fine-grained realism, Kling Video O1 is a fantastic new option.

Jump into your next scene and give it a try — it’s available today for all users.

***

#### Tue Nov 25, 2025

<img alt="A middle aged woman sitting at a cafe" />

**FLUX.2 is now live in Automat Studio!**

The latest model from Black Forest Labs brings sharper detail, richer texture, and incredibly photorealistic lighting to your shots. Generate characters, locations, and scenes with a whole new level of clarity and style.

Try FLUX.2 today: [https://www.automatstudio.com/](https://www.automatstudio.com/)

***

Tue, Nov 25, 2025

<iframe title="Character Studio Tutorial" />

**Introducing Character Studio**

This is where your characters come to life. Define their name, age, and appearance, generate their look, and choose a voice that shapes their personality. Start creating the characters who inhabit your world.

Try it today: [https://www.automatstudio.com/](https://www.automatstudio.com/)

***

#### Thu, Nov 20, 2025

<img alt="A young woman drawing cinematic storyboards" />

**Google Nano Banana 2 (Pro) is now live in Automat Studio**

We have added Google’s new Nano Banana Pro model to Automat Studio. This model brings major improvements for filmmakers and creators, including stronger character consistency, and clean text rendering inside images. Nano Banana Pro is now available in Character Studio, Location Studio, and Shot Studio. Try it today and explore a new level of visual quality inside Automat Studio.

Try it today: [https://www.automatstudio.com/](https://www.automatstudio.com/)

***

#### Tue, November 18, 2025

<img alt="A young woman drawing cinematic storyboards" />

**Automat Studio doesn’t generate ideas—you do.**

The artist is always in control. Behind the scenes, our vision is made possible by new generative AI technologies. But we firmly believe AI’s purpose is not to replace human creativity.

Creativity is the responsibility of the artist, and great filmmaking will always need a human spark. Automat Studio is a new tool to help bring your vision to life—your characters, your locations, your scenes. The story is always yours.

Learn more at [Automat Studio](https://automatstudio.com).

***

#### Tue, November 11, 2025

**How to Visualize Your Film With Automat Studio (In 90 Seconds)**

We’ve added so many new features to Automat Studio that we had to remake our 90-second tutorial just a few months after the first one came out! Check out the all-new “Automat Studio in 90 Seconds.” It’ll show you how to create characters, locations, scenes, and shots to bring your film to life with Automat Studio.

<iframe title="How to Visualize Your Film With Automat Studio (In 90 Seconds)" />

Try it now: [Automat Studio](https://automatstudio.com).

***

#### Thu, June 26, 2025

<video>
  <source type="video/mp4" />

  Your browser does not support the video tag.
</video>

**Kling 2.1 Arrives in Automat Studio**

We’re excited to announce that Kling 2.1—the newest version of Kuaishou’s acclaimed text-to-video model—is now live in Shot Studio.

Kling 2.1 elevates motion fidelity, frame-to-frame coherence, and cinematic lighting, turning a single prompt into fluid, story-ready footage in seconds. Build sweeping establishing shots, fast-cut action beats, or quiet character moments while preserving visual continuity across every clip.

Try Kling 2.1 now: [Automat Studio](https://automatstudio.com).

***

#### Tue, June 24, 2025

<img alt="Paris Bistro" />

**Automat Studio Features OpenAI GPT Image 1**

We’re excited to announce that Automat Studio now includes OpenAI’s GPT Image 1 model — a powerful addition to our AI image generation lineup.

Known for its versatility and creativity, GPT Image 1 delivers high-quality visuals ideal for character concepts, location design, and cinematic world-building. Whether you need photorealism, stylization, or conceptual artwork, GPT Image 1 helps bring your ideas to life.

It’s available now in Character Studio and Location Studio.

Explore it today: [Automat Studio](https://automatstudio.com).

***

#### Thu, June 19, 2025

<video>
  <source type="video/mp4" />

  Your browser does not support the video tag.
</video>

**Runway Gen-4 Has Arrived in Automat Studio.**

We’re excited to announce that Runway’s Gen-4 model is now integrated into Character Studio, Location Studio, and Shot Studio.

As a recognized leader in AI-powered video, Runway has consistently pushed the boundaries of what’s possible in narrative filmmaking. Gen-4 continues that tradition — delivering expressive characters, stylized environments, and cinematic video clips with a handcrafted feel.

Whether you’re building a mood, testing a concept, or crafting full sequences, Gen-4 helps bring your story to life with coherence and tone across every visual layer.

Available now at [Automat Studio](https://automatstudio.com).

***

#### Tue, June 17, 2025

<img alt="Four Charater Headshots" />

**Automat Studio Features Google Imagen 4.**

We’re excited to share that Automat Studio includes Google Imagen 4, one of the most advanced image generation models available today.

You can use Imagen 4 to create rich, detailed, photorealistic character portraits and location visuals directly within Character Studio and Location Studio.

Imagen 4 produces stunningly realistic faces, environments, and textures—perfect for cinematic concept work, storyboards, or production design inspiration.

Try it now at [Automat Studio](https://automatstudio.com).

***

#### Thu, June 12, 2025

<img alt="Two filmmakers watching a video monitor" />

**Automat Studio is structured around the traditional filmmaking process.**

We’ve designed Automat Studio to reflect the roles of an actual film production. Character Studio is your casting department. Location Studio handles locations and set decoration. Aesthetics covers cinematography. Shot Creator lets the director compose the frame. More departments are coming soon.

Filmmaking has always been a collaborative craft. With Automat Studio, we’re preserving that structure — while making it easier to experiment, iterate, and bring your vision to life.

If you know how films are made, Automat Studio will feel familiar. It’s built on the same creative foundations.

Explore it today at [Automat Studio](https://automatstudio.com).

***

#### Tue, June 10, 2025

<img alt="A young woman drawing cinematic storyboards" />

**Automat Studio doesn’t generate ideas—you do.**

The artist is always in control. Behind the scenes, our vision is made possible by new generative AI technologies. But we firmly believe AI’s purpose is not to replace human creativity.

Creativity is the responsibility of the artist, and great filmmaking will always need a human spark. Automat Studio is a new tool to help bring your vision to life—your characters, your locations, your scenes. The story is always yours.

Learn more at [Automat Studio](https://automatstudio.com).

***

#### Thu, June 5, 2025

<img alt="Automat Studio Logo" />

**Welcome to Automat Studio.**

Hi everyone, I’m Paul.

Over the past few months, I’ve been building a new tool for filmmakers called Automat Studio.

It’s designed to help you visualize your film before you shoot. You can create characters, define locations, plan scenes, and generate images or video to explore your concept early in the process.

Whether you’re a writer, director, or anyone working on a film, Automat Studio offers a way to develop and refine your ideas.

The public beta is now live. If you’re curious, you’re welcome to join the waitlist and take a look.

[https://www.automatstudio.com/](https://www.automatstudio.com/)


# How to Build an AI Storyboard and Shot List
Source: https://docs.automatstudio.com/guides/ai-storyboards-and-shot-lists

Build a shot list and a visual storyboard for your film with AI — one shot per setup, each with a lens, a shot type and a generated frame you can show a crew.

In Automat Studio a storyboard and a shot list are the same object viewed two
ways. You break each [scene](/tutorial/create-a-scene) into shots, give every
shot a lens, a shot type and a description of the action, and generate a
[first frame](/tutorial/shot-studio/generate-first-frame) for it. The list of
shots is your shot list; the generated frames, in order, are your storyboard.
Both stay live, so changing a shot updates the board.

Because the frames are generated from characters and locations you have already
defined, the board shows your actual cast in your actual sets — not generic
sketches.

## Step 1: List the setups before you generate

Work scene by scene and decide the coverage first, in plain language: a wide to
establish, a single on each character, an insert on the object that matters.
Create one shot per setup. It is much cheaper to reorganise a list of empty
shots than to regenerate frames you did not need.

## Step 2: Give each shot its grammar

Open a shot in [Shot Studio](/tutorial/shot-studio/index) and set:

* **Shot type** — the framing, from wide through to close-up.
* **Lens** — which changes how compressed and how intimate the frame reads.
* **Description** — what actually happens in this framing, written for this
  shot rather than for the scene as a whole.

Then `@`-mention the [characters](/tutorial/character-studio) and
[props](/tutorial/prop-studio) that appear, so the frame is populated with your
established assets. See
[mentioning assets](/tutorial/shot-studio/mention-assets).

## Step 3: Generate the frames in Draft first

Every generation has a **Draft / Final** quality switch. A storyboard is
exactly what Draft is for: you want to see blocking, composition and continuity
across twenty frames, not a finished image of one. Board the whole scene in
Draft, look at it as a sequence, fix what does not read, and only push the
frames you will actually use up to Final.

This is also the cheapest way to work. Generations are paid for in credits, new
accounts get 500 free, and failed or cancelled generations are never charged.

## Step 4: Read the board as a sequence

With the scene's frames generated, look at them in order and check the things a
board is meant to catch:

* Does the coverage cut together, or are you missing a connecting shot?
* Is screen direction consistent from shot to shot?
* Does the same character look like the same person throughout? If not, see
  [keeping characters consistent](/guides/consistent-ai-characters).
* Does the [aesthetic](/tutorial/aesthetics) hold across the scene?

Reordering, duplicating and deleting shots is cheap, so treat the first pass as
a draft of the sequence rather than a commitment.

## Step 5: Turn the board into the film

The difference between this and a paper storyboard is that the board is the
film's first stage rather than a document you set aside. When a frame is right,
[turn it into video](/tutorial/shot-studio/generate-video) — image-to-video
starts from the frame you approved, so the shot you boarded is the shot you
get. Then add
[dialogue and lip sync](/tutorial/shot-studio/dialogue-lipsync),
[sound effects](/tutorial/shot-studio/sfx-sound-effects) and
[visual effects](/tutorial/shot-studio/vfx-visual-effects) per shot.

## Sharing the board

You can share your project with a link so collaborators can see the sequence —
shares from free accounts carry a watermark. Independent plans and above can
[export](/tutorial/share-and-export) the frames as image files to drop into a
deck or a PDF, and the Hollywood plan can export an XML project for Final Cut
Pro or Premiere.

## One honest limitation

Lens and shot type steer a generation rather than dictate it — the docs are
explicit that shot attributes may not be reflected precisely in the image, a
current limitation of the underlying models. For boarding this matters less
than it sounds, because what you are checking is blocking, sequence and
continuity. When an exact framing is essential, iterate on the shot description
rather than relying on the lens setting alone.

## Next steps

* [How to make a film with AI](/guides/how-to-make-a-film-with-ai)
* [Turn a script into video](/guides/turn-a-script-into-video)
* [Shot Studio tutorial](/tutorial/shot-studio/index)


# How to Keep AI Characters Consistent Across Shots
Source: https://docs.automatstudio.com/guides/consistent-ai-characters

Why AI characters drift between shots, and how defining them once as reusable assets keeps the same face, build and wardrobe in every scene of your film.

AI characters drift between shots because each generation starts from scratch:
if the character only exists in the prompt text, the model re-invents them
every time. The fix is to define the character once as a reusable asset with a
generated reference image, then pull that same asset into every shot instead of
re-describing them. In Automat Studio that is what
[Character Studio](/tutorial/character-studio) and `@`-mentions are for.

## Why drift happens

A text prompt is a fresh instruction. "A woman in her thirties with dark hair"
describes a space of thousands of different people, and a video model picks
somewhere in that space each time you ask. Add a new adjective in shot four and
you have moved to a different part of the space — so shot four shows a
different person.

Nothing in a single prompt carries an identity forward. Consistency has to come
from reusing a fixed reference, not from writing a better description.

## Step 1: Define the character once, in detail

Create the character before you generate any shots. Give them a name, an age,
and a description that pins down what the model would otherwise invent:

* **Face and build** — hair colour, length and style; eye colour; height;
  build; skin tone; age markers.
* **Distinctive features** — scars, glasses, freckles, tattoos, the things that
  make a face recognisable.
* **Default wardrobe** — what they wear when nothing else is specified.
* **Bearing** — how they hold themselves, which affects how they read on
  screen.

Vague descriptions are the single biggest cause of drift. "Tall with an
angular face, cropped bleached hair and a chipped front tooth" survives
regeneration; "attractive young man" does not.

You also select the character's **voice** here, which is what gets used later
when you generate their [dialogue](/tutorial/shot-studio/dialogue-lipsync) — so
the voice stays consistent too.

## Step 2: Lock a reference image you are happy with

Generate the character's image and keep regenerating until one is right, then
make it the active generation. That image is the visual anchor for the
character across the film, so it is worth spending credits to get it right
before you shoot anything. Use the Draft quality setting while you are
exploring and Final once you are close.

## Step 3: Mention the character in every shot

In [Shot Studio](/tutorial/shot-studio/index), type `@` and pick the character
to pull them into the shot rather than describing them again. The same works
for [props](/tutorial/prop-studio), [locations](/tutorial/location-studio) and
[wardrobe](/tutorial/wardrobe-studio) — see
[mentioning assets](/tutorial/shot-studio/mention-assets).

This is the habit that matters. Any shot where you type the description by hand
instead of mentioning the asset is a shot that can drift.

## Step 4: Let the aesthetic carry the look

Character consistency is only half of visual continuity — the other half is
grade, lighting and texture. An [aesthetic](/tutorial/aesthetics) applied at
project level keeps those consistent across every generation, so shots cut
together as though one crew shot them. Six aesthetics are built in, and
Independent and Hollywood plans can build custom ones from reference images.

## Step 5: Anchor motion to a still

For a shot where the character's face really matters, generate a
[first frame](/tutorial/shot-studio/generate-first-frame) you are happy with
and then [turn that frame into video](/tutorial/shot-studio/generate-video).
Image-to-video starts from your approved still, so the face in the moving shot
is the face you signed off — which is more reliable than generating video
directly from text.

If you need a specific performance, you can go further and upload your own
filmed take with
[performance to video](/tutorial/shot-studio/performance): the motion comes
from your real performance and the AI restyles it.

## Wardrobe and props drift too

The same rules apply. Define
[wardrobe](/tutorial/wardrobe-studio) and [props](/tutorial/prop-studio) as
assets, link wardrobe to a character where relevant, and mention them per shot.
A consistent face in an inconsistent jacket still breaks the cut.

## What to expect

This approach gets you a recognisably consistent character across a film, which
is what narrative work needs. It is not frame-perfect identity lock: current
models still vary, and shot attributes such as lens and shot type steer a
generation rather than dictate it. Budget for iteration on the shots where the
face is large in frame, and use the reference-anchored image-to-video path for
those.

## Next steps

* [How to make a film with AI](/guides/how-to-make-a-film-with-ai)
* [Turn a script into video](/guides/turn-a-script-into-video)
* [Character Studio tutorial](/tutorial/character-studio)


# How to Make a Film With AI
Source: https://docs.automatstudio.com/guides/how-to-make-a-film-with-ai

A step-by-step guide to making a short film with AI — from script to finished shots — using a real production workflow instead of a single prompt.

To make a film with AI in Automat Studio you work the way a real production
does, in five steps: create a project, define your characters and locations as
reusable assets, break the story into scenes, direct each shot individually,
then assemble and export the result. The AI generates the images, video and
audio; you make every creative decision. You can do the whole thing in a
browser, and a new account starts with 500 free credits and no credit card.

The reason to work this way rather than typing one long prompt into a video
model is consistency. A prompt gives you an isolated clip; a production gives
you the same character, in the same wardrobe, in the same location, across
every shot in the film.

## Step 1: Create a project

A project is one film. You give it a name, a synopsis, and an
[aesthetic](/tutorial/aesthetics) — the visual style that will be applied
across everything you generate, so the whole film looks like it was shot by
one crew. Six aesthetics are built in; Independent and Hollywood plans can
also create custom ones from reference images.

Start here: [Create a Project](/tutorial/create-a-project).

## Step 2: Cast your characters and build your world

This is the step that separates a film from a pile of clips. Each
[character](/tutorial/character-studio) is defined once — name, age, a detailed
physical description, and a voice — and then reused everywhere. Do the same for
[locations](/tutorial/location-studio) (interior or exterior, time of day),
[props](/tutorial/prop-studio) and [wardrobe](/tutorial/wardrobe-studio).

Write these descriptions the way a screenwriter or a production designer would:
specific about hair, build, distinctive features, clothing and light. The
detail you put in here is what the AI reuses in every later shot.

## Step 3: Break the story into scenes

A [scene](/tutorial/create-a-scene) is a story moment: one location, plus the
characters and props that appear in it, plus a time of day and an optional
synopsis. You need at least one location before you can create a scene.

Scenes are how the film gets its structure. They are also how you keep track of
what you have and have not shot yet.

## Step 4: Direct each shot

Inside a scene you create shots, and each one opens
[Shot Studio](/tutorial/shot-studio/index). For a shot you choose the lens and
the shot type, describe the action, and `@`-mention the characters and props
that appear in it so the AI uses the versions you already designed.

From there you can:

* **Generate a first frame** — the hero still that anchors the shot, then
  [turn it into video](/tutorial/shot-studio/generate-video).
* **Generate video directly**, skipping the still, when you already know what
  you want.
* **Add [dialogue and lip sync](/tutorial/shot-studio/dialogue-lipsync)** in
  your character's selected voice.
* **Add [sound effects](/tutorial/shot-studio/sfx-sound-effects)** and
  [visual effects](/tutorial/shot-studio/vfx-visual-effects).

Every generation has a Draft/Final quality switch, so you can block out a whole
scene cheaply and only spend on final quality once the shot works.

Be aware of one honest limitation: shot attributes such as lens and shot type
are not always reflected precisely in the generated image. They steer the
result rather than dictate it, and this improves as the underlying models do.

## Step 5: Watch, share and export

When the shots are generated you can watch the assembled film in the app and
share it with a link — shares from free accounts carry a watermark. Independent
plans and above can [export](/tutorial/share-and-export) the underlying image,
video and audio files, and the Hollywood plan can export an XML project for
Final Cut Pro or Premiere to finish the edit in your own timeline.

## What it costs

Everything you generate is paid for with credits, and every new account gets
500 of them free with no credit card required. After that you can buy a
one-time credit pack from $10 — purchased credits never expire — or subscribe
to a monthly plan from $10, each of which includes monthly credits worth 20%
more than the plan price. Browsing, organising a project and failed generations
never cost credits. See [pricing](https://automatstudio.com/pricing) and the
[credits guide](/support/usage) for the details.

## Which AI models does it use?

You are not locked to one model. Automat Studio brings the leading ones
together in a single studio — Google Veo and Imagen, OpenAI Sora and GPT-Image,
Runway Gen-4 and Aleph, Kling, Seedream, FLUX and more, with ElevenLabs voices
for dialogue — and you pick the right one per shot. The
[model reference](/models/index) lists every supported model with its credit
price.

## Next steps

* [Start the full tutorial](/tutorial/index)
* [Turn an existing script into video](/guides/turn-a-script-into-video)
* [Keep characters consistent across shots](/guides/consistent-ai-characters)


# How to Turn a Script Into Video With AI
Source: https://docs.automatstudio.com/guides/turn-a-script-into-video

Take an existing screenplay or script and turn it into video, scene by scene and shot by shot, keeping your characters and locations consistent throughout.

To turn a script into video with AI in Automat Studio, you map the script onto
the app's structure rather than pasting it into a prompt box: each character in
your script becomes a reusable character asset, each slugline becomes a
location and a scene, and each beat within a scene becomes a shot you direct
individually. Then you generate the shots and assemble them into the film.

That mapping is the whole technique. A screenplay is already broken into the
units a production uses, so most of the structuring work is done before you
start.

## Map the script onto the app

| In your script | In Automat Studio |
| - | - |
| Character introduced in action lines | A [character](/tutorial/character-studio) — name, age, physical description, voice |
| `INT. KITCHEN – NIGHT` | A [location](/tutorial/location-studio) (interior/exterior + time of day), then a [scene](/tutorial/create-a-scene) |
| Objects the action depends on | [Props](/tutorial/prop-studio) and [wardrobe](/tutorial/wardrobe-studio) |
| A beat or camera setup within the scene | A shot in [Shot Studio](/tutorial/shot-studio/index) |
| A line of dialogue | [Dialogue and lip sync](/tutorial/shot-studio/dialogue-lipsync) on that shot |

## Step 1: Pull the characters out first

Go through the script and create every speaking character before you generate
anything. Use the description from your action lines and expand it — hair,
build, age, distinctive features, wardrobe, and the way they carry themselves.
Assign each one a voice; that is the voice used later when you generate their
dialogue.

Doing this first is what makes the rest consistent. The character is generated
once and then reused, so the person in scene 12 is the person from scene 1.

## Step 2: Turn sluglines into locations and scenes

Every distinct slugline becomes a location, described the way a production
designer would describe the set. Then create a scene for each one: pick the
location, mark it interior or exterior, set the time of day, and add the
characters and props that appear. You need at least one location before you can
create a scene.

If your script returns to the same location at a different time of day, reuse
the location — the time of day lives on the scene, not the location.

## Step 3: Break each scene into shots

This is where you direct. A page of screenplay is rarely one shot: it is a
wide, a couple of singles, maybe an insert. For each shot, choose the lens and
shot type, describe the action in that specific framing, and `@`-mention the
characters and props involved so the generation uses the assets you already
built.

Then generate — either a
[first frame](/tutorial/shot-studio/generate-first-frame) that you turn into
[video](/tutorial/shot-studio/generate-video), or
[video directly](/tutorial/shot-studio/generate-video-direct) when the shot is
simple. Use the Draft quality setting to block out the scene, then regenerate
the keepers at Final.

## Step 4: Add the dialogue

With the shot's video in place, generate the character's line in their assigned
voice and apply lip sync. You can also add
[sound effects](/tutorial/shot-studio/sfx-sound-effects) and
[visual effects](/tutorial/shot-studio/vfx-visual-effects) per shot.

If you have a filmed performance you want to drive the shot with — your own
take of the action — you can upload it and use
[performance to video](/tutorial/shot-studio/performance) instead, which
restyles your real performance rather than generating the motion from scratch.

## Step 5: Assemble and export

Watch the cut in the app and share it with a link. Independent plans and above
can [export](/tutorial/share-and-export) the image, video and audio files, and
the Hollywood plan exports an XML project for Final Cut Pro or Premiere — which
is usually how a script-derived piece gets finished, since you will want real
editorial control over timing.

## Practical notes

* **Work scene by scene, not front to back.** Getting one scene fully right
  teaches you how to describe the rest.
* **Costs are per generation, in credits.** New accounts get 500 free credits
  with no credit card; after that, packs start at $10 and plans at $10/month.
  Failed generations are not charged, and a running generation can be cancelled for a full refund.
* **Shot attributes steer rather than dictate.** Lens and shot type influence
  the generation but are not always reflected precisely — expect to iterate on
  the description for a specific framing.

## Next steps

* [How to make a film with AI](/guides/how-to-make-a-film-with-ai)
* [Keep characters consistent across shots](/guides/consistent-ai-characters)
* [Build a storyboard and shot list](/guides/ai-storyboards-and-shot-lists)


# Introduction
Source: https://docs.automatstudio.com/index

Welcome to the Automat Studio documentation

## Welcome to Automat Studio

Automat Studio is an AI filmmaking studio that lets you make your film. Build characters, design locations, structure scenes, and generate shots—all within an integrated workflow that maintains visual consistency across your entire production.

This documentation will guide you through every aspect of using Automat Studio, from your first project to advanced techniques and troubleshooting.

***

## Get Started

<CardGroup>
  <Card title="Tutorial" icon="book-open" href="/tutorial/index">
    Follow the complete production workflow from project setup to final rendered shots.
  </Card>

  <Card title="Aesthetics" icon="palette" href="/tutorial/aesthetics">
    Explore visual styles and create custom aesthetics for your productions.
  </Card>
</CardGroup>

***

## Help & Support

<CardGroup>
  <Card title="Frequently Asked Questions" icon="circle-question" href="/support/faq">
    Find answers to common questions about Automat Studio features and workflows.
  </Card>

  <Card title="Credits & Subscriptions" icon="gauge" href="/support/usage">
    Understand how credits work and what each subscription plan includes.
  </Card>

  <Card title="Troubleshooting" icon="triangle-exclamation" href="/support/troubleshooting">
    Resolve common issues and get help when things don't go as expected.
  </Card>
</CardGroup>


# AI Content Disclaimer
Source: https://docs.automatstudio.com/legal/ai-content

How AI-generated content works in Automat Studio and who owns it

**Effective Date:** May 24, 2025\
**Last Updated:** May 29, 2025

Automat Studio LLC provides tools for generating AI-based images and videos from user-submitted text and story ideas. The following disclaimers apply to all AI-generated content:

***

## 1. Fictional Nature

All generated media is intended for creative storytelling purposes.\
Any resemblance to real persons, places, or events is purely coincidental and unintended.

***

## 2. No Human Oversight

AI-generated output is automated and not reviewed by humans before delivery. As a result:

* Content may contain errors or anomalies.
* Quality and appropriateness may vary.
* Generated content may reflect the limitations or biases of the underlying models.

***

## 3. User Responsibility

You are solely responsible for how you use, distribute, or publish the content you create on our platform.\
You must not use the Service to generate or share illegal, harmful, or offensive content.

We reserve the right to review and remove any content that violates these guidelines.

***

## 4. Ownership and Rights

As stated in our Terms of Service, you retain full ownership of all AI-generated content you create using the platform.

***

## 5. Limitations of Use

We do not support the use of AI content generated from our platform in any context involving:

* Misrepresentation or deception
* Sensitive, defamatory, or harmful material
* Deepfakes of real individuals without consent

***

## 6. Training and Model Use

We do not use your content to train any AI models, either directly or indirectly.\
If this policy changes in the future, we will request your explicit consent.


# Cookie Policy
Source: https://docs.automatstudio.com/legal/cookies

How Automat Studio uses cookies and how to control them

**Effective Date:** May 24, 2025\
**Last Updated:** May 29, 2025

This Cookie Policy explains how Automat Studio LLC (“we”, “us”, “our”) uses cookies and similar technologies on our website.

***

## 1. What Are Cookies?

Cookies are small text files stored on your device when you visit a website. They help websites remember your activity and preferences.

***

## 2. Cookies We Use

We aim to minimize tracking and only use essential, privacy-focused analytics tools:

* **Vercel Analytics** – used to understand aggregate traffic patterns.
* **Plausible Analytics** – cookie-free, GDPR-compliant analytics.

These analytics tools are operated by third parties and subject to their own privacy practices.

We do not use third-party advertising cookies or fingerprinting.

***

## 3. Your Consent

By continuing to use our website, you consent to our use of cookies as described in this policy.

***

## 4. How to Control Cookies

Although we use minimal cookies, you can control or delete cookies through your browser settings.\
Visit [www.allaboutcookies.org](https://www.allaboutcookies.org) for more information.

***

## 5. Changes to This Policy

We may update this policy as we add features or tools.

***

## 6. More Information

Please review our [Privacy Policy](/legal/privacy) for more information on how we handle your data.

***

## 7. Contact

For any cookie-related questions, contact: [hello@automatstudio.com](mailto:hello@automatstudio.com)


# Privacy Policy
Source: https://docs.automatstudio.com/legal/privacy

How Automat Studio collects, uses, and protects your data

**Effective Date:** May 24, 2025\
**Last Updated:** June 25, 2026

Automat Studio LLC (“we,” “our,” or “us”) is committed to protecting your privacy. This Privacy Policy explains how we collect, use, and store your information when you use our website and services (the “Service”).

By using the Service, you agree to the collection and use of your information in accordance with this policy.

***

## 1. Information We Collect

### a. Account Information

* Email address and authentication provider (e.g., Google) used for login.
* No passwords are stored by us directly (handled by our auth provider).

### b. User Input

* Names, descriptions, and other fictional story content entered into the platform.
* AI-generated images and videos based on your prompts.

### c. Analytics & Cookies

* We use Vercel Analytics and Plausible Analytics to gather non-personal, aggregated usage data (e.g., page views, session duration).
* These two tools do not track personal identifiers or set tracking cookies.
* We also use Google Ads conversion tracking (the Google tag) to measure the effectiveness of our advertising. With your consent, this may set advertising cookies and process identifiers (including a hashed email — see Section 3). Non-essential storage is governed by Google Consent Mode and is denied by default until you accept cookies; you can decline at any time via the cookie banner.

***

## 2. How We Use Your Information

* To provide and improve the Service.
* To store and maintain your projects and content.
* To understand general user behavior and platform usage trends.

We do not use your data to train AI models.

***

## 3. Data Sharing

We do not sell, rent, or share your personal data with third parties, except:

* With service providers (e.g., hosting, authentication) under confidentiality agreements.
* If required by law or to protect rights and safety.

We also integrate with third-party AI generation APIs (such as OpenAI, RunwayML, and FAL.ai) to deliver features of the Service. Any data sent to those services is subject to their privacy policies.

### Third-Party Service Providers and Data Sharing

We share customer data with third parties to perform services on our behalf. We obtain consent for such sharing where required by law or any applicable Google policies governing personalized ads and/or user consent, including Google's EU user consent policy.

**Google Ads Customer Match:** We may share customer data with Google for Customer Match campaigns. The data files we upload will only be used to:

* Match your customers to Google accounts
* Ensure our Customer Match campaigns comply with Google's policies

Under the Google Ads Data Processing Terms, which are incorporated into the Google Ads terms of service, Google acts as a "processor" of the personal data we may share with them for Customer Match. To comply with the General Data Protection Regulation (GDPR), we ensure that our account settings maintain up-to-date contact information for our primary contact and data protection officer and/or EU representative (if applicable). Any notices under the Google Ads Data Processing Terms will be sent to the primary contact.

**Google Ads Enhanced Conversions:** When you complete a purchase, we send Google a hashed (irreversible, SHA-256) version of your email address through Google's Enhanced Conversions feature. This lets Google measure the effectiveness of our advertising by matching your purchase to a prior ad interaction in a privacy-preserving way. The hashing happens in your browser before any data is sent, so we never transmit your email address to Google in plain text. This data is shared only where you have consented and is processed in accordance with the Google Ads Data Processing Terms and Google's [Privacy Policy](https://policies.google.com/privacy).

***

## 4. Data Retention

We retain your data as long as your account is active or as needed to provide you the Service. You may request deletion of your data at any time by contacting us.

You may also request a copy of your stored data for export.

***

## 5. Data Storage & Security

We use industry-standard methods (e.g., Supabase, Vercel) to store and secure your data.\
However, no system is completely secure, so we cannot guarantee absolute security.

***

## 6. Children’s Privacy

Our Service is not intended for children under 13. We do not knowingly collect data from children under 13.

***

## 7. Your Rights

If you would like to request deletion or a copy of your data or account, you may contact us at [hello@automatstudio.com](mailto:hello@automatstudio.com).

***

## 8. International Users

We comply with applicable international privacy laws, including the General Data Protection Regulation (GDPR) for users in the EU/EEA. If you are accessing the Service from outside the United States, you do so at your own discretion, and your data may be transferred to and processed in the United States.

***

## 9. Data Subject Rights (EU/EEA Users)

If you are located in the EU or EEA, you may have additional rights under the GDPR, including the right to access, correct, delete, or restrict the use of your personal data.

***

## 10. Consent

By using the Service, you consent to:

* Our use of cookies and analytics tools as described in this policy
* Our sharing of your data with third-party service providers, including Google, for the purposes described in this policy
* The processing of your personal data by third parties acting as processors on our behalf, including Google for Customer Match campaigns, in accordance with applicable data processing terms and where required by law

***

## 11. Changes to This Policy

We may update this Privacy Policy. Changes will be posted on this page and take effect immediately.

***

## 12. Contact

**Automat Studio LLC**\
447 Broadway, 2nd Floor #1715
New York, NY 10013
[hello@automatstudio.com](mailto:hello@automatstudio.com)


# Terms of Service
Source: https://docs.automatstudio.com/legal/terms

The terms governing your use of Automat Studio

**Effective Date:** May 24, 2025\
**Last Updated:** Feb 11, 2026

Welcome to Automat Studio LLC (“we”, “our”, or “us”). These Terms of Service (“Terms”) govern your access to and use of our website and services (collectively, the “Service”), including the ability to generate AI images and videos, and create fictional characters, locations, and story elements.

By using the Service, you agree to these Terms. If you do not agree, do not use the Service.

***

## 1. Eligibility

You must be at least 13 years old to use the Service. If you are under 18, you must have a parent or legal guardian’s permission.

***

## 2. Your Content

You may enter descriptive text and upload images to create fictional characters, locations, and story elements.

* **Ownership**: You retain ownership of original content you create using the Service, including AI-generated images and videos, subject to these Terms.
* **Your Responsibility**: You represent and warrant that any content you submit or upload:
  * Is original to you or that you have the necessary rights and permissions to use it;
  * Does not infringe or violate the intellectual property, publicity, privacy, or other rights of any third party;
  * Does not depict or attempt to recreate the likeness of any real individual without their express written permission.
* **No Celebrity or Public Figure Likenesses**: You may not upload images of, or attempt to generate content depicting, any real person, including actors, musicians, athletes, politicians, influencers, public figures, or other recognizable individuals.
* **License to Operate the Service**: You grant us a limited, non-exclusive license to host, store, process, and display your content solely for the purpose of operating and improving the Service.
* **Storage**: We store your projects and data for your continued access, but we do not train AI models on your content.

You are solely responsible for how you use, publish, distribute, or otherwise exploit any content generated through the Service.

***

## 3. Privacy

Your projects are private by default.\
We use analytics tools (Vercel and Plausible) to understand general usage patterns, but we do not collect personally identifiable tracking data unless explicitly provided by you.

For more information, please review our [Privacy Policy](/legal/privacy).

***

## 4. Prohibited Use

You agree not to:

* Use the Service for unlawful, harmful, or offensive purposes.
* Submit prompts or names that include explicit sexual content, nudity, violence, or illegal substances. Examples of blocked terms include pornography, torture, rape, and drug trafficking. See our documentation for the full policy.
* Attempt to reverse engineer, copy, or misuse the Service.
* Submit prompts or names that infringe on the rights of others.
* Upload, submit, or generate content that depicts or attempts to replicate the likeness, voice, or identity of any real person without their documented consent.
* Upload copyrighted photographs or images unless you are the rights holder or have explicit permission to use them.
* Use the Service to create deepfakes, impersonations, or misleading representations of real individuals.

**Warning:** Any attempt to create sexually explicit content will result in immediate account termination without warning.

***

## 5. Likeness and Right of Publicity

You acknowledge that individuals have legal rights in their name, image, likeness, voice, and identity (commonly known as “right of publicity” rights). You agree not to use the Service to create, upload, generate, or distribute content that violates these rights.

Automat Studio LLC reserves the right to remove any content that we believe may infringe upon a person’s publicity or privacy rights, even if the individual is a public figure.

***

## 6. Third-Party AI Services

We use external APIs and AI providers, including but not limited to OpenAI, RunwayML, and FAL.ai, to generate media. These services are operated by third parties, and we do not control their models, outputs, availability, or policies.

We make no representations or warranties regarding the accuracy, legality, appropriateness, or ownership status of content generated by third-party AI services. You acknowledge that AI-generated output may be inaccurate, incomplete, infringing, or otherwise legally problematic, and you assume all risk associated with your use of such content.

Your use of the Service may also be subject to the terms and policies of those third-party providers.

***

## 7. Account Termination

We may suspend or terminate your account or access to the Service at any time, without notice, if we believe you have violated these Terms or if we discontinue the Service.

***

## 8. Availability

We do our best to maintain the availability of the Service but do not guarantee uptime or uninterrupted access. The Service may be unavailable during maintenance or due to technical issues beyond our control.

***

## 9. Subscriptions and Credits

The Service uses credits to pay for generation. Usage is based on generation activity, not time spent in the Service.

* **What spends credits:** Generating character images, location images, first frames, and videos.
* **What does not spend credits:** Browsing, organizing projects, preparing assets, and failed generations (including content filter failures).
* **Plan credits:** Each subscription tier includes a monthly grant of credits. Unused plan credits do not carry over to the following month; they reset with each billing cycle and expire if the subscription is cancelled.
* **Purchased credits:** Credit packs purchased separately do not expire and remain available until used. Plan credits are consumed before purchased credits.
* **Upgrades:** If you upgrade your subscription mid-cycle, the charge will be prorated for the remaining time in your billing period, and you will receive additional plan credits for the remainder of the month.
* **Downgrades:** Downgrades are not prorated and take effect at the end of your current billing cycle. You will retain access to your current plan until the end of the billing period.

For more details on subscription plans and credits, see our [Credits & Subscriptions guide](/support/usage).

***

## 10. Refunds

Except where required by law, all payments are final and non-refundable.

***

## 11. Feedback

If you submit feedback, suggestions, or ideas to us, you grant us a non-exclusive, royalty-free license to use them for any purpose, including improving the Service.

***

## 12. Modifications

We may update these Terms occasionally. If we make material changes, we will notify you via the Service or email.

***

## 13. Disclaimers

The Service is provided "as is" and "as available," without warranties of any kind, whether express or implied. To the maximum extent permitted by law, Automat Studio LLC disclaims all warranties, including but not limited to warranties of merchantability, fitness for a particular purpose, non-infringement, and any warranties arising out of course of dealing or usage of trade.

We do not guarantee that the Service will be uninterrupted, secure, or error-free. AI-generated image and video outputs are probabilistic and may contain inaccuracies, distortions, artifacts, or unexpected results. We do not warrant that generated content will be legally compliant, non-infringing, or suitable for your intended use.

***

## 14. Limitation of Liability

To the maximum extent permitted by law, Automat Studio LLC shall not be liable for any indirect, incidental, special, consequential, or punitive damages, including loss of profits, revenue, data, or goodwill, arising out of or related to your use of the Service.

In no event shall Automat Studio LLC’s total liability to you for all claims arising out of or relating to the Service exceed the total amount you paid to us in the twelve (12) months preceding the event giving rise to the claim.

Some jurisdictions do not allow certain limitations of liability, so some of the above limitations may not apply to you.

***

## 15. Governing Law

These Terms are governed by the laws of New York, United States, without regard to conflict of law principles.

***

## 16. Indemnification

You agree to defend, indemnify, and hold harmless Automat Studio LLC, its officers, employees, contractors, and affiliates from and against any claims, damages, obligations, losses, liabilities, costs, or expenses (including reasonable attorneys’ fees) arising from:

* Your use of the Service;
* Your violation of these Terms;
* Your infringement of any intellectual property, publicity, privacy, or other rights of any person or entity.

***

## 17. Copyright Complaints (DMCA)

Automat Studio LLC respects intellectual property rights and complies with the Digital Millennium Copyright Act (17 U.S.C. §512).

If you believe that content hosted on the Service infringes your copyright, you must submit a written notice to [hello@automatstudio.com](mailto:hello@automatstudio.com) that includes:

* Identification of the copyrighted work claimed to have been infringed;
* Identification of the material that is claimed to be infringing and sufficient information for us to locate it;
* Your name, address, telephone number, and email address;
* A statement that you have a good faith belief that use of the material is not authorized by the copyright owner, its agent, or the law;
* A statement that the information in the notice is accurate, and under penalty of perjury, that you are authorized to act on behalf of the copyright owner;
* Your physical or electronic signature.

We may remove or disable access to allegedly infringing material and may terminate repeat infringers at our discretion.

***

## 18. Contact

For questions, contact us at: [hello@automatstudio.com](mailto:hello@automatstudio.com)


# Audio-to-Video
Source: https://docs.automatstudio.com/models/audio-to-video

Sync character lip movement and expressions to dialogue audio

Audio-to-Video models analyze dialogue audio and generate synchronized lip movements, facial expressions, and subtle head motions for characters. In Automat Studio, these models power the Dialogue and Lipsync workflow, automatically aligning character animation to voice performances.

<Note>
  These models are used in the Dialogue and Lipsync tab of Shot Studio to automatically sync character lip movements and expressions to dialogue audio.
</Note>

## When to Use

* **Automatic Lipsync** — Sync character mouths to dialogue without manual animation
* **Expression Matching** — Generate facial expressions that match dialogue emotion
* **Character Animation** — Add subtle head movements and reactions
* **Post-Production Sync** — Align existing character shots with dialogue tracks

## Recommended Models

These models consistently deliver the best results for lipsync:

### Sync Lipsync 2 Pro

* **Credits:** 63
* **Rating:** ⭐⭐⭐⭐ (4/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~100 seconds
* **Best for:** Professional-quality lip synchronization (Default model)

### Sync React-1

* **Credits:** 125
* **Rating:** ⭐⭐⭐⭐ (4/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~266 seconds
* **Best for:** Advanced reactive expressions and head movements

## Supported Models

No additional supported models are currently available for this workflow.

## Tips for Best Results

1. **Clean Audio** — Use isolated dialogue without background music for best accuracy
2. **Tag Speakers** — Specify which character is speaking to improve sync accuracy
3. **Include Emotion Cues** — Mention the emotion or tone in prompts for better expression matching
4. **Preview Playback** — Review synced results against the waveform to confirm timing
5. **High Quality Source** — Clear, well-recorded dialogue produces better sync results
6. **Multiple Passes** — Adjust and regenerate if initial sync isn't perfect

> ⚠️ **Important:** High-quality, clean audio without background music leads to the most accurate lip sync results.

***

<CardGroup>
  <Card title="Dialogue and Lipsync" icon="microphone" href="/tutorial/shot-studio/dialogue-lipsync">
    Learn how to sync dialogue to your shots
  </Card>

  <Card title="Shot Studio" icon="film" href="/tutorial/shot-studio/index">
    Complete guide to shot creation and enhancement
  </Card>
</CardGroup>


# Image-to-Image
Source: https://docs.automatstudio.com/models/image-to-image

Edit and refine single images with AI-powered transformations

Image-to-Image models transform and edit single images based on text prompts. In Automat Studio, these models enable quick edits to character portraits, location images, and shot frames without starting from scratch.

<Note>
  These models are used for editing and refining individual images across Character Studio, Location Studio, and Shot Studio.
</Note>

## When to Use

* **Quick Edits** — Make targeted changes to existing images (clothing, lighting, details)
* **Style Adjustments** — Modify the visual style or mood of an image
* **Detail Refinement** — Enhance or correct specific areas of generated images
* **Iterative Improvement** — Build on successful generations with targeted edits

## Recommended Models

These models consistently deliver the best results for image editing:

### Google Nano Banana 2.1

* **Credits:** 8
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** Google
* **Best for:** Precise edits that change only what you ask, with mark-the-region editing (Default model)

### Google Nano Banana 2

* **Credits:** 11
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** Google
* **Avg. Duration:** \~18 seconds
* **Best for:** Premium quality image editing with exceptional detail

### Bytedance Seedream V5 Lite Edit

* **Credits:** 6
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** Cost-effective, high-quality edits

### Bytedance Seedream V5 Pro Edit

* **Credits:** 11
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** Region-precise edits that keep the rest of the frame intact

### Google Nano Banana Lite Edit

* **Credits:** 6
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** Fast, cheap multi-turn local edits

### Google Nano Banana Pro

* **Credits:** 21
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** Google
* **Avg. Duration:** \~16 seconds
* **Best for:** Premium quality edits with exceptional detail

### Kling 3.0 Image

* **Credits:** 5
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~53 seconds
* **Best for:** High-quality, affordable editing

### OpenAI GPT Image 2

* **Credits:** 27
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** OpenAI
* **Avg. Duration:** \~165 seconds
* **Best for:** Demanding edits with excellent prompt comprehension; superseded by GPT Image 2.5 for speed and cost

### OpenAI GPT Image 2.5 Flare

* **Credits:** 9
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** OpenAI
* **Best for:** Precise edits that change only what you ask for while preserving the subject; far faster and cheaper than GPT Image 2

### OpenAI GPT Image 2.5 Sunburst

* **Credits:** 9
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** OpenAI
* **Best for:** Premium edit precision across complex subjects and backgrounds; takes longer than Flare

### Seedream 4.5

* **Credits:** 6
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~30 seconds
* **Best for:** High-quality edits with excellent detail preservation

### FLUX.2 \[max]

* **Credits:** 11
* **Rating:** ⭐⭐⭐⭐ (4/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~35 seconds
* **Best for:** Higher quality FLUX editing for demanding edits

### FLUX.2 \[pro]

* **Credits:** 5
* **Rating:** ⭐⭐⭐⭐ (4/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~16 seconds
* **Best for:** Fast, cost-effective editing with good quality

### Google Nano Banana

* **Credits:** 6
* **Rating:** ⭐⭐⭐⭐ (4/5)
* **Provider:** Google
* **Avg. Duration:** \~17 seconds
* **Best for:** Fast, high-quality edits at low cost

## Supported Models

These models are available but may have varying quality or processing times:

### Reve

* **Credits:** 6
* **Rating:** ⭐⭐⭐ (3/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~27 seconds
* **Use when:** Creative reinterpretations are desired

## Tips for Best Results

1. **Be Specific** — Clearly describe what you want to change and what should stay the same
2. **Start Small** — Make targeted edits rather than trying to change everything at once
3. **Preserve Context** — Mention elements you want to keep unchanged in your prompt
4. **Iterate** — Build on successful edits with additional refinements
5. **Match Style** — Ensure edits align with your project's aesthetic

***

<CardGroup>
  <Card title="Character Studio" icon="users" href="/tutorial/character-studio">
    Edit character images in Character Studio
  </Card>

  <Card title="Location Studio" icon="location-dot" href="/tutorial/location-studio">
    Refine location images in Location Studio
  </Card>
</CardGroup>


# Image-to-Video
Source: https://docs.automatstudio.com/models/image-to-video

AI models for animating still images into motion clips

Image-to-Video models transform static images into animated video clips. In Automat Studio, these models power the "Generate Video" workflow in Shot Studio, bringing your first frames to life with realistic motion, camera movements, and temporal consistency.

<Note>
  These models are used to generate videos from first frame images in the Generate Video tab of Shot Studio.
</Note>

<Info>
  Video credits scale with clip **duration** and whether **audio** is generated. The figures below show the range across supported durations; the "(with audio)" figure applies when audio generation is enabled.
</Info>

## When to Use

* **Shot Animation** — Animate your approved first frame into a complete shot
* **Motion Generation** — Add camera movement, character motion, and environmental dynamics
* **Narrative Progression** — Create temporal continuity between static frames
* **Shot Completion** — Transform still compositions into dynamic cinematic moments

## Recommended Models

These models consistently deliver the best results for video generation:

### Google Veo 3.1

* **Credits:** 120–240 (240–480 with audio)
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~258 seconds
* **Best for:** Highest quality motion generation

### Happy Horse 1.1

* **Credits:** 63–315
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~148 seconds
* **Best for:** The latest Happy Horse with improved motion quality

### LTX-2.5 Fast

* **Credits:** 81–270
* **Rating:** ⭐⭐⭐⭐ (4/5)
* **Provider:** fal.ai
* **Best for:** Fast clips with synchronized native audio and lip-sync, up to 20 seconds

### LTX-2.5 Pro

* **Credits:** 108–180
* **Rating:** ⭐⭐⭐⭐ (4/5)
* **Provider:** fal.ai
* **Best for:** Higher-fidelity LTX generation with synchronized native audio and lip-sync

### LTX-2.3 Fast

* **Credits:** 36–120
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~29 seconds
* **Best for:** Rapid, high-quality generation

### LTX-2.3 Pro

* **Credits:** 54–90
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~75 seconds
* **Best for:** High-quality LTX motion generation

<Warning>
  Seedance 2.0 and 2.5 reject images that look like real people — an upstream content policy that can also catch AI-generated photoreal faces. A rejected request fails with a content-safety message and the credits are refunded. For shots built around photoreal human characters, pick another model (Kling, Veo, MiniMax H3) and keep Seedance for stylized or non-human shots.
</Warning>

### Seedance 2.5

* **Credits:** 278–2081
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** Native clips up to 30 seconds in a single pass — no drift or multi-clip stitching

### Seedance 2.0

* **Credits:** 182–681
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** Maximum quality motion generation for final shots

### Seedance 2.0 Fast

* **Credits:** 146–545
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~164 seconds
* **Best for:** High-quality motion with faster turnaround

### Seedance 2.0 Mini

* **Credits:** 90–341
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~211 seconds
* **Best for:** Affordable, high-quality Seedance generation

### Kling O3 Pro

* **Credits:** 51–252 (63–315 with audio)
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** First/last-frame transitions and multi-shot generation at top Kling quality

### Kling O3 Standard

* **Credits:** 38–189 (51–252 with audio)
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** First/last-frame transitions at a lower cost

### Gemini Omni Flash 1.1

* **Credits:** 45–150
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** Physics-grounded motion with native synchronized audio, sharper than Omni Flash 1.0 (always included)

### Gemini Omni Flash

* **Credits:** 59–195
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** Physics-grounded motion with native synchronized audio (always included)

### MiniMax H3

* **Credits:** 60–180
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** Takes up to 15 seconds with native stereo audio (always included) at a low per-second cost

### MiniMax H3 Max

* **Credits:** 60–180
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** The same price and native stereo audio as MiniMax H3, generated several times faster

### MiniMax H3 Max Turbo

* **Credits:** 18–90
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** H3 Max motion and native audio at half the price, generated even faster

### Vidu Q4

* **Credits:** 44–228
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** Up to 16 seconds with native audio and dialogue, from a single frame (always included)

### FLUX 3 Video

* **Credits:** 128–510
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** Black Forest Labs' frontier video model: up to 20 seconds with optional audio

### Wan 3.0

* **Credits:** 30–450
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** Clips from 2 up to 30 seconds with native audio, at the same per-second cost as Wan 2.7

### Wan 3.0 Prime

* **Credits:** 42–630
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** Alibaba's premium tier — richer motion and cinematic detail on clips up to 30 seconds

### Wan 2.7

* **Credits:** 30–225
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** Smooth, coherent motion at an affordable per-second cost

### Google Veo 3.1 Fast

* **Credits:** 60–120 (90–180 with audio)
* **Rating:** ⭐⭐⭐⭐ (4/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~79 seconds
* **Best for:** Quick iterations with excellent motion quality (Default model)

### Google Veo 3 Fast

* **Credits:** 75 (113 with audio)
* **Rating:** ⭐⭐⭐⭐ (4/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~137 seconds
* **Best for:** Fast, high-quality results

### Kling 3.0 Standard

* **Credits:** 126 (189 with audio)
* **Rating:** ⭐⭐⭐⭐ (4/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~81 seconds
* **Best for:** Smooth motion with excellent temporal consistency

### Seedance 1.5

* **Credits:** 20 (39 with audio)
* **Rating:** ⭐⭐⭐⭐ (4/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~100 seconds
* **Best for:** High-quality motion at an affordable cost

## Supported Models

These models are available but may have varying quality, speed, or motion consistency:

### Kling 2.6 \[Pro]

* **Credits:** 53 (105 with audio)
* **Rating:** ⭐⭐⭐ (3/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~94 seconds
* **Use when:** Smooth, high-quality Kling motion is needed

### Kling 3.0 Pro

* **Credits:** 168 (252 with audio)
* **Rating:** ⭐⭐⭐ (3/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~112 seconds
* **Use when:** Premium Kling quality is the priority

## Tips for Best Results

1. **Start with Strong First Frame** — A clear, well-composed first frame yields better motion
2. **Adjust Motion Settings** — Experiment with motion strength and camera path before generating
3. **Be Patient** — Video generation typically takes 2-3 minutes; you can work on other shots meanwhile
4. **Try Multiple Attempts** — Each generation is unique; regenerate if motion isn't right
5. **Review and Iterate** — Watch the result, adjust settings, and regenerate until satisfied

> ⚠️ **Processing Time:** Video generation is computationally intensive. Most models take 1-4 minutes. You can navigate away from the page while rendering—Automat processes in the background.

***

<CardGroup>
  <Card title="Generate Video" icon="video" href="/tutorial/shot-studio/generate-video">
    Learn how to use these models in Shot Studio
  </Card>

  <Card title="Shot Studio" icon="film" href="/tutorial/shot-studio/index">
    Complete guide to shot creation workflows
  </Card>
</CardGroup>


# Introduction
Source: https://docs.automatstudio.com/models/index

Complete guide to all AI models available in Automat Studio

Automat Studio integrates with cutting-edge AI models from leading providers including Google, OpenAI, Runway, and others. Each model type is optimized for specific creative workflows in the filmmaking process.

<Tip>
  In the app, the model picker (**Click here to change** under the generate button) shows the same one-line summary for every model as these pages, with the **Draft** and **Final** models listed first and the credit cost on every card.
</Tip>

## Model Categories

<CardGroup>
  <Card title="Text-to-Image" icon="image" href="/models/text-to-image">
    Generate character and location images from text descriptions
  </Card>

  <Card title="Image-to-Image" icon="pen-to-square" href="/models/image-to-image">
    Edit and refine single images with targeted changes
  </Card>

  <Card title="Multiple Image-to-Image" icon="images" href="/models/multiple-image-to-image">
    Refine images using reference images and prompts
  </Card>

  <Card title="Reference-to-Image" icon="clone" href="/models/reference-to-image">
    Generate assets from reference images and a prompt
  </Card>

  <Card title="Image-to-Video" icon="video" href="/models/image-to-video">
    Animate still images into motion clips
  </Card>

  <Card title="Multiple Image-to-Video" icon="film" href="/models/multiple-image-to-video">
    Create video from multiple reference images
  </Card>

  <Card title="Performance-to-Video" icon="masks-theater" href="/models/performance-to-video">
    Re-render a real filmed performance in a new world
  </Card>

  <Card title="Video-to-Video" icon="wand-magic-sparkles" href="/models/video-to-video">
    Enhance and style existing video clips
  </Card>

  <Card title="Text-to-Audio" icon="microphone" href="/models/text-to-audio">
    Generate dialogue and narration from text
  </Card>

  <Card title="Voice Models" icon="user" href="/models/voice-models">
    Browse available AI voice models for audio generation
  </Card>

  <Card title="Audio-to-Video" icon="lips" href="/models/audio-to-video">
    Sync lip movement and expressions to audio
  </Card>

  <Card title="Video-to-Audio" icon="waveform" href="/models/video-to-audio">
    Generate sound effects from video footage
  </Card>
</CardGroup>

***

## Model Quality Tiers

### Recommended Models

**Recommended models** are tested and optimized for Automat Studio workflows. They consistently deliver the best results for narrative filmmaking with reliable quality, faster processing times, and better integration with our platform.

### Supported Models

**Supported models** are available for use but may vary in quality, consistency, or processing speed. They can be useful for:

* Specific artistic styles or experimental workflows
* Testing different approaches to generation
* Cost optimization for certain use cases

In Automat Studio, the model in use is shown just below each generate button — click **Click here to change** to see and select from all available models for each workflow.

***

## Understanding Model Costs

Each model uses **credits** based on:

* Generation complexity
* Output resolution and quality
* Processing time required
* Provider pricing

For video models, credits also scale with the clip **duration** and whether **audio** is generated, so those pages show a credit range rather than a single figure.

Check individual model pages for specific credit costs. You can monitor your credit usage in your account dashboard.

***

## Model Providers

Automat Studio works with leading AI providers:

* **Google** — Imagen, Veo, and other cutting-edge models
* **OpenAI** — GPT Image and Sora video models
* **Runway** — Gen-4 Image and Video generation tools
* **fal.ai** — Curated selection of open and proprietary models

Each provider brings unique strengths to different creative tasks.


# Multiple Image-to-Image
Source: https://docs.automatstudio.com/models/multiple-image-to-image

Refine and generate images using reference images and prompts

Multiple Image-to-Image models use reference images combined with text prompts to generate refined or stylized versions. In Automat Studio, these models power the "Generate First Frame" workflow in Shot Studio, allowing you to iterate on shot compositions using previous frames or reference images.

<Note>
  These models are used in the Generate First Frame tab of Shot Studio to iterate on shot compositions.
</Note>

## When to Use

* **First Frame Generation** — Create the hero still for shots using shot descriptions and visual references
* **Style Consistency** — Maintain visual coherence across shots by referencing previous frames
* **Iterative Refinement** — Build upon successful generations to perfect compositions
* **Shot Matching** — Align new shots with existing visual style from the project

## Recommended Models

These models consistently deliver the best results for shot frame generation:

### Google Nano Banana 2.1

* **Credits:** 12
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** Google
* **Best for:** Consistent characters and objects composed from up to 14 references (Default model)

### Google Nano Banana 2

* **Credits:** 11
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** Google
* **Avg. Duration:** \~26 seconds
* **Best for:** Premium quality editing with exceptional detail

### Bytedance Seedream V5 Lite Edit

* **Credits:** 6
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** Cost-effective, high-quality refinement

### FLUX.2 \[pro]

* **Credits:** 5
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~33 seconds
* **Best for:** Fast, high-quality editing with excellent prompt adherence

### Google Nano Banana

* **Credits:** 6
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** Google
* **Avg. Duration:** \~16 seconds
* **Best for:** Quick iterations and experimentation

### Google Nano Banana Pro

* **Credits:** 21
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** Google
* **Avg. Duration:** \~35 seconds
* **Best for:** Premium quality image editing with exceptional detail

### OpenAI GPT Image 2

* **Credits:** 29
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** OpenAI
* **Best for:** Complex compositions with excellent prompt comprehension; superseded by GPT Image 2.5 for speed and cost

### OpenAI GPT Image 2.5 Flare

* **Credits:** 12
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** OpenAI
* **Best for:** Composing characters, wardrobe and locations into one frame with strong subject preservation, at speed

### OpenAI GPT Image 2.5 Sunburst

* **Credits:** 12
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** OpenAI
* **Best for:** The most controlled multi-reference compositions; slower than Flare

### Runway Gen-4 Image Turbo (References)

* **Credits:** 3
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** Runway
* **Avg. Duration:** \~26 seconds
* **Best for:** Fast, high-quality iterations with excellent reference understanding

### Seedream 4.5

* **Credits:** 6
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~53 seconds
* **Best for:** High-quality editing with excellent detail

### Qwen Image Edit Plus

* **Credits:** 5
* **Rating:** ⭐⭐⭐⭐ (4/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~20 seconds
* **Best for:** Cost-effective editing and refinement

### Reve Remix

* **Credits:** 6
* **Rating:** ⭐⭐⭐⭐ (4/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~38 seconds
* **Best for:** Creative reinterpretations with high quality

### Runway Gen-4 Image (References)

* **Credits:** 12
* **Rating:** ⭐⭐⭐⭐ (4/5)
* **Provider:** Runway
* **Avg. Duration:** \~39 seconds
* **Best for:** Highest quality results when you need maximum fidelity

### Vidu Reference-to-Image

* **Credits:** 15
* **Rating:** ⭐⭐⭐⭐ (4/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~49 seconds
* **Best for:** Working with video reference frames

## Supported Models

These models are available but may have varying quality or processing times:

### OmniGen v2

* **Credits:** 45
* **Rating:** ⭐⭐⭐ (3/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~116 seconds
* **Use when:** Exploring alternative generation methods

## Tips for Best Results

1. **Use Clear References** — Provide high-quality reference images that clearly show what you want
2. **Combine with Prompts** — Text descriptions enhance reference image understanding
3. **Iterate Systematically** — Build on successful generations rather than starting from scratch
4. **Maintain Shot History** — Save multiple variations to compare and choose the best
5. **Match Camera Settings** — Reference frames work best when shot attributes (lens, angle) are similar

***

<CardGroup>
  <Card title="Generate First Frame" icon="image" href="/tutorial/shot-studio/generate-first-frame">
    Learn how to use these models in Shot Studio
  </Card>

  <Card title="Shot Studio" icon="video" href="/tutorial/shot-studio/index">
    Complete guide to shot creation and generation
  </Card>
</CardGroup>


# Multiple Image-to-Video
Source: https://docs.automatstudio.com/models/multiple-image-to-video

Generate video from multiple reference images or storyboards

Multiple Image-to-Video models create motion clips from a sequence of reference images or storyboard frames. In Automat Studio, these models power the "Generate Video Direct" workflow, allowing you to skip the first frame step and generate motion directly from visual references.

<Note>
  These models are used in the Generate Video Direct tab of Shot Studio to create motion from multiple reference images or storyboard frames.
</Note>

<Info>
  Video credits scale with clip **duration** and whether **audio** is generated. Where a range is shown, it reflects the supported durations for that model.
</Info>

## When to Use

* **Generate Video Direct** — Skip first frame generation and create video from uploaded references
* **Storyboard to Video** — Transform storyboard sequences into animated clips
* **Reference-Based Animation** — Generate motion that respects specific visual references
* **Rapid Prototyping** — Quickly test motion concepts from existing images

## Recommended Models

These models consistently deliver the best results for multi-image video generation:

### Google Veo 3.1 (Ingredients)

* **Credits:** 150
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~53 seconds
* **Best for:** Highest quality motion from reference sequences

### Happy Horse 1.1

* **Credits:** 63–315
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~199 seconds
* **Best for:** The latest Happy Horse with improved motion quality

### Kling Video O1 Reference

* **Credits:** 84
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~146 seconds
* **Best for:** Sharp motion with excellent reference understanding (Default model)

<Warning>
  Seedance 2.0 and 2.5 reject images that look like real people — an upstream content policy that can also catch AI-generated photoreal faces. A rejected request fails with a content-safety message and the credits are refunded. For shots built around photoreal human characters, pick another model (Kling, Veo, MiniMax H3) and keep Seedance for stylized or non-human shots.
</Warning>

### Seedance 2.5

* **Credits:** 278–2081
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** Native clips up to 30 seconds in a single pass with locked character, set, and palette consistency

### Seedance 2.0

* **Credits:** 182–681
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** Maximum quality reference-based generation

### Seedance 2.0 Fast

* **Credits:** 146–545
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~202 seconds
* **Best for:** High-quality motion with faster turnaround

### Seedance 2.0 Mini

* **Credits:** 90–341
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~219 seconds
* **Best for:** Affordable, high-quality Seedance generation

### Kling O3 Pro Reference

* **Credits:** 51–252
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** Stable character identity and object detail across reference-based scenes

### Kling O3 Standard Reference

* **Credits:** 38–189
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** Reference-based scenes at a lower cost

### Gemini Omni Flash 1.1 Reference

* **Credits:** 45–150
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** Multimodal reference-to-video with native synchronized audio and consistent characters (always included)

### Vidu Q4 Reference

* **Credits:** 44–228
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** Up to 12 references kept consistent across a 16-second take, with native audio (always included)

### Gemini Omni Flash Reference

* **Credits:** 59–195
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** Multimodal reference-to-video with native synchronized audio (always included)

### MiniMax H3

* **Credits:** 60–180
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** Holding up to 9 subject and style references consistent, with native stereo audio (always included)

### MiniMax H3 Max

* **Credits:** 60–180
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** The same references and native stereo audio as MiniMax H3, generated several times faster

### Wan 3.0

* **Credits:** 30–450
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** Up to 10 reference images driving clips from 2 to 30 seconds

### Wan 3.0 Prime

* **Credits:** 42–630
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** Premium reference fidelity — character identity and style held across long clips

## Tips for Best Results

1. **Consistent Reference Images** — Use images that share similar style, lighting, and composition
2. **Clear Story Progression** — Reference images should show logical progression of action or movement
3. **Match Shot Attributes** — Ensure references align with your shot's camera settings and movement
4. **Quality References** — Higher quality input images produce better motion results
5. **Experiment with Sequence** — Try different numbers of reference images (2-5 works well)

***

<CardGroup>
  <Card title="Generate Video Direct" icon="bolt" href="/tutorial/shot-studio/generate-video-direct">
    Learn how to use these models in Shot Studio
  </Card>

  <Card title="Shot Studio" icon="film" href="/tutorial/shot-studio/index">
    Complete guide to shot creation workflows
  </Card>
</CardGroup>


# Performance-to-Video
Source: https://docs.automatstudio.com/models/performance-to-video

Generate video from a real filmed performance plus a styled first frame

Performance-to-Video models re-render a real filmed take into a new world: your uploaded performance drives the motion — movement, timing, blocking, expressions — while your styled first frame drives the look. In Automat Studio, these models power [Performance mode](/tutorial/shot-studio/performance) in Shot Studio.

<Note>
  Takes must be 2–30 seconds; output duration is matched to your take. (The Seedance 2.0 models support takes up to 15 seconds — models that can't cover your take are hidden from the model picker.) The take's original audio — your actors' real voices and on-set sound — can be carried into the generated video with the "Keep my take's audio" toggle.
</Note>

<Info>
  Video credits scale with clip **duration** (2–30 seconds, matched to your take). Keeping the take's audio costs nothing extra.
</Info>

## Recommended Models

### Kling 3.0 Motion

* **Credits:** 39–567
* **Rating:** ⭐⭐⭐⭐ (4/5)
* **Provider:** fal.ai
* **Best for:** High-quality performance transfer at the best price (Default model)

### Kling 3.0 Motion Pro

* **Credits:** 51–756
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** Maximum fidelity for complex movement and expressive performances

### MiniMax H3 Max

* **Credits:** 95–506
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** Fast performance transfer on takes up to 15 seconds, with generated audio

<Note>
  MiniMax H3 Max works a little differently from the Kling Motion models: it always generates its own audio, so the "Keep my take's audio" toggle has no effect on it, and its shortest output is 5 seconds — a take under 5 seconds produces a 5-second clip. It accepts takes up to 15 seconds.
</Note>

## Supported Models

<Warning>
  The Seedance models (2.0 and 2.5) reject reference footage containing likenesses of real people (an upstream content policy — including AI-restyled frames of real people). Use them only for stylized or non-human takes; for filmed performances of real actors, use the Kling Motion models above. With Seedance, the audio toggle generates AI audio rather than keeping your take's sound.
</Warning>

### Seedance 2.5

* **Credits:** 167–2496
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** The strongest Seedance quality and consistency for stylized/non-human takes (up to 30s)

### Seedance 2.0

* **Credits:** 92–681
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** Maximum-quality reference generation for stylized/non-human takes

### Seedance 2.0 Fast

* **Credits:** 74–545
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** Faster Seedance generation for stylized/non-human takes

### Seedance 2.0 Mini

* **Credits:** 45–341
* **Rating:** ⭐⭐⭐⭐ (4/5)
* **Provider:** fal.ai
* **Best for:** Affordable Seedance generation for stylized/non-human takes


# Reference-to-Image
Source: https://docs.automatstudio.com/models/reference-to-image

Generate assets from reference images and a text prompt

Reference-to-Image models generate a new image from one or more **reference images** combined with a text prompt. In Automat Studio, these models power the "from reference" workflow used when creating assets — building a character, location, prop, or wardrobe item that matches the look of images you provide.

<Note>
  These models are used in Character Studio, Location Studio, Prop Studio, and Wardrobe Studio when you generate an asset from reference images.
</Note>

<Warning>
  **Your reference is not copied.** Automat generates a *new* image from it — a studio portrait for a character, a cinematic wide shot for a location, a product-style shot for a prop, and a ghost-mannequin shot with nobody wearing it for a wardrobe item. Expect a close likeness, not an exact reproduction of the file you uploaded. That is what makes the asset reusable: it has to be re-posed, re-lit and placed into shots that don't exist in your original photo.
</Warning>

## When to Use

* **Asset from References** — Create a character, location, prop, or wardrobe item that matches uploaded reference images
* **Consistency** — Keep a recurring asset visually consistent across your project
* **Likeness Matching** — Generate variations that preserve the identity or style of your references
* **Concept Realization** — Turn moodboard or inspiration images into a usable asset

## Recommended Models

These models consistently deliver the best results for reference-based generation:

### Google Nano Banana 2.1

* **Credits:** 8
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** Google
* **Best for:** Strong likeness from a reference at a lower cost than Nano Banana 2 (Default model)

### Google Nano Banana 2

* **Credits:** 11
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** Google
* **Avg. Duration:** \~18 seconds
* **Best for:** Premium quality with strong reference adherence

### OpenAI GPT Image 2

* **Credits:** 27
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** OpenAI
* **Best for:** Excellent prompt comprehension for demanding reference matching

### OpenAI GPT Image 2.5 Flare

* **Credits:** 9
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** OpenAI
* **Best for:** Strong likeness preservation from a reference photo, quickly

### OpenAI GPT Image 2.5 Sunburst

* **Credits:** 9
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** OpenAI
* **Best for:** Tighter control when the reference has a complex subject or background; slower than Flare

### Google Nano Banana

* **Credits:** 6
* **Rating:** ⭐⭐⭐⭐ (4/5)
* **Provider:** Google
* **Avg. Duration:** \~12 seconds
* **Best for:** Fast, high-quality generation at low cost

### Kling 3.0 Image

* **Credits:** 5
* **Rating:** ⭐⭐⭐⭐ (4/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~38 seconds
* **Best for:** Affordable, high-quality reference-based generation

## Upload Requirements

One reference image per asset, in JPG, PNG, or WebP, up to **10MB**.

## Tips for Best Results

1. **Use Clear References** — Provide high-quality, well-lit reference images that clearly show the subject
2. **Combine with Prompts** — A descriptive prompt sharpens what the model takes from your references
3. **Match the Goal** — Use references that reflect the angle, framing, and style you want in the result
4. **Iterate** — Generate multiple versions and keep the ones that best match your references
5. **Keep Assets Consistent** — Reuse the same references when generating variations of a recurring asset
6. **Push Likeness with Final** — For a real face, a sharp, well-lit, front-on photo on **Final** quality holds the likeness noticeably better than Draft. Lock the reference image you are happy with before you shoot any shots

***

<CardGroup>
  <Card title="Character Studio" icon="users" href="/tutorial/character-studio">
    Create characters from reference images
  </Card>

  <Card title="Wardrobe Studio" icon="shirt" href="/tutorial/wardrobe-studio">
    Build wardrobe items from references
  </Card>
</CardGroup>


# Text-to-Audio
Source: https://docs.automatstudio.com/models/text-to-audio

Generate dialogue and narration from text scripts

Text-to-Audio models transform written scripts into spoken dialogue or narration. In Automat Studio, these models power the dialogue generation workflow, allowing you to create voice performances directly from your scripts without recording sessions.

<Note>
  These models are used to generate dialogue audio from text scripts for your shots.
</Note>

## When to Use

* **Dialogue Generation** — Create character dialogue from script text
* **Narration** — Generate voiceover or narration tracks
* **Rapid Prototyping** — Test dialogue pacing and delivery before final recording
* **Multiple Takes** — Generate variations of the same line for director review

## Recommended Models

These models consistently deliver the best results for voice generation:

### ElevenLabs v3

* **Credits:** 2
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~35 seconds
* **Best for:** Natural-sounding dialogue with excellent voice quality and emotion

## Supported Models

No additional supported models are currently available for this workflow.

## Tips for Best Results

1. **Break Long Scripts** — Process scenes separately to manage file sizes
2. **Specify Emotion and Tone** — Include delivery cues in your prompts (pacing, emotion, intensity)
3. **Generate Multiple Takes** — Create several versions to offer director options
4. **Match Character Voices** — Use consistent voice profiles for each character
5. **Preview Before Using** — Listen to generated audio before integrating into shots
6. **Clean Audio Preferred** — Isolated dialogue works better than text with sound effects

***

<CardGroup>
  <Card title="Dialogue and Lipsync" icon="microphone" href="/tutorial/shot-studio/dialogue-lipsync">
    Learn how to use dialogue in your shots
  </Card>

  <Card title="Shot Studio" icon="film" href="/tutorial/shot-studio/index">
    Complete guide to shot creation workflows
  </Card>
</CardGroup>


# Text-to-Image
Source: https://docs.automatstudio.com/models/text-to-image

AI models for generating character and location images from text descriptions

Text-to-Image models transform written descriptions into high-quality images. In Automat Studio, these models power character and location generation, allowing you to create consistent visual assets for your film by describing them in natural language.

<Note>
  These models are used to generate character and location images in Character Studio and Location Studio.
</Note>

## When to Use

* **Character Creation** — Generate character portraits from descriptions of appearance, age, clothing, and personality
* **Location Design** — Create environments and settings from architectural and atmospheric descriptions
* **Rapid Prototyping** — Quickly visualize concepts before committing to detailed production

## Recommended Models

These models consistently deliver the best results for narrative filmmaking:

### Google Nano Banana 2.1

* **Credits:** 8
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** Google
* **Best for:** Sharper detail and stronger prompt adherence than Nano Banana 2, at a lower cost (Default model)

### Google Nano Banana 2

* **Credits:** 11
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** Google
* **Avg. Duration:** \~22 seconds
* **Best for:** Premium quality with exceptional detail and prompt adherence

### FLUX.2 \[pro]

* **Credits:** 5
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~27 seconds
* **Best for:** Fast, high-quality generation with excellent prompt adherence

### Google Nano Banana

* **Credits:** 6
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** Google
* **Avg. Duration:** \~15 seconds
* **Best for:** Quick, high-quality iterations and experimentation

### Google Nano Banana Pro

* **Credits:** 21
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** Google
* **Avg. Duration:** \~26 seconds
* **Best for:** Premium quality images with exceptional detail

### Kling 3.0 Image

* **Credits:** 5
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~51 seconds
* **Best for:** High-quality, affordable generation

### OpenAI GPT Image 2

* **Credits:** 26
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** OpenAI
* **Avg. Duration:** \~148 seconds
* **Best for:** Excellent prompt comprehension; superseded by GPT Image 2.5 for speed and cost

### OpenAI GPT Image 2.5 Flare

* **Credits:** 8
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** OpenAI
* **Best for:** The newest GPT Image model: sharper detail and stronger prompt adherence than GPT Image 2 at a fraction of the credits and time

### OpenAI GPT Image 2.5 Sunburst

* **Credits:** 8
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** OpenAI
* **Best for:** The same quality as Flare with tighter control on demanding work; slower to generate

### Seedream 4.5

* **Credits:** 6
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~29 seconds
* **Best for:** High-quality generation with excellent detail

### Bytedance Seedream V5 Lite

* **Credits:** 6
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~53 seconds
* **Best for:** Cost-effective, high-quality generation

### Bytedance Seedream V5 Pro

* **Credits:** 11
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** Flagship Seedream quality with precise control over dense, detailed scenes

### Google Nano Banana 2 Lite

* **Credits:** 6
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** Sub-2-second draft generations at minimal cost

### Google Imagen 4

* **Credits:** 8
* **Rating:** ⭐⭐⭐⭐ (4/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~16 seconds
* **Best for:** High-quality images with exceptional detail and photorealism

### Hunyuan Image 3.0

* **Credits:** 15
* **Rating:** ⭐⭐⭐⭐ (4/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~95 seconds
* **Best for:** Complex scenes with multiple elements

### Runway Gen-4 Image

* **Credits:** 12
* **Rating:** ⭐⭐⭐⭐ (4/5)
* **Provider:** Runway
* **Avg. Duration:** \~30 seconds
* **Best for:** Cinematic quality with excellent composition and lighting

## Supported Models

These models are available but may have varying quality or consistency:

### FLUX 1.1 \[pro] (ultra)

* **Credits:** 9
* **Rating:** ⭐⭐⭐ (3/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~14 seconds
* **Use when:** Detailed, artistic images with excellent prompt adherence are required

### MiniMax (Hailuo AI) Image-01

* **Credits:** 2
* **Rating:** ⭐⭐⭐ (3/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~27 seconds
* **Use when:** Budget is the primary concern and you need quick iterations

### Recraft V3

* **Credits:** 6
* **Rating:** ⭐⭐⭐ (3/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~15 seconds
* **Use when:** Cost-effective high-quality generation is needed

### Reve

* **Credits:** 6
* **Rating:** ⭐⭐⭐ (3/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~15 seconds
* **Use when:** High-quality results with creative flexibility are needed

## Tips for Best Results

1. **Be Specific** — Include details about appearance, lighting, mood, and composition in your descriptions
2. **Iterate** — Generate multiple versions to find the best result
3. **Use Reference Terms** — Mention artistic styles, time periods, or visual references
4. **Save Successful Prompts** — Keep track of descriptions that work well for consistency
5. **Combine with Aesthetics** — Project aesthetics can enhance and unify results across different models

***

<CardGroup>
  <Card title="Character Studio" icon="users" href="/tutorial/character-studio">
    Learn how to create characters using these models
  </Card>

  <Card title="Location Studio" icon="location-dot" href="/tutorial/location-studio">
    Create locations with text-to-image models
  </Card>
</CardGroup>


# Video-to-Audio
Source: https://docs.automatstudio.com/models/video-to-audio

Generate sound effects and ambient audio from video footage

Video-to-Audio models analyze video footage and generate matching sound effects, ambient audio, and foley sounds. In Automat Studio, these models power the SFX (Sound Effects) workflow, automatically creating soundscapes that sync to on-screen action.

<Note>
  These models are used in the SFX tab of Shot Studio to automatically generate sound effects that match the visual action in your shots.
</Note>

## When to Use

* **Automatic SFX Generation** — Create sound effects that match on-screen action
* **Ambient Sound Design** — Generate environmental audio beds for scenes
* **Foley Generation** — Add spot effects like footsteps, impacts, and movement sounds
* **Rapid Audio Prototyping** — Quickly test sound design concepts

## Recommended Models

These models consistently deliver the best results for sound effect generation:

### Mirelo SFX v1.5

* **Credits:** 8
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~14 seconds
* **Best for:** Highest quality sound effect generation (Default model)

## Supported Models

No additional supported models are currently available for this workflow.

## Tips for Best Results

1. **Quality Source Video** — Upload high-quality renders for better audio analysis
2. **Consistent Frame Rate** — Stable frame rates help with accurate timing
3. **Action Clarity** — Clear, visible action produces better-matched sound effects
4. **Mix Layers** — Use separate stems (FX, ambience, impacts) for flexible mixing
5. **Adjust Levels** — Balance SFX volume so dialogue remains clear
6. **Export Stems** — Save layered stems for use in your DAW or editor

> 💡 **Tip:** You can always replace or layer additional audio tracks later if you need to make adjustments or add director-requested changes.

***

<CardGroup>
  <Card title="SFX (Sound Effects)" icon="volume-high" href="/tutorial/shot-studio/sfx-sound-effects">
    Learn how to add sound effects to your shots
  </Card>

  <Card title="Shot Studio" icon="film" href="/tutorial/shot-studio/index">
    Complete guide to shot creation and audio
  </Card>
</CardGroup>


# Video-to-Video
Source: https://docs.automatstudio.com/models/video-to-video

Enhance and refine existing video clips with visual effects and styling

Video-to-Video models transform existing video clips by applying visual effects, style transfers, or quality enhancements. In Automat Studio, these models power the VFX (Visual Effects) workflow in Shot Studio, allowing you to layer cinematic effects onto your rendered shots without regenerating from scratch.

<Note>
  These models are used in the VFX tab of Shot Studio to enhance existing video clips with visual effects, lighting adjustments, and stylized treatments.
</Note>

## When to Use

* **VFX Application** — Add visual effects like lighting tweaks, particles, or overlays
* **Style Transfer** — Apply different visual styles to existing footage
* **Quality Enhancement** — Improve or refine rendered shots
* **Non-Destructive Editing** — Modify shots without losing original work
* **Color Grading** — Adjust color palettes and tonal ranges

## Recommended Models

These models consistently deliver the best results for video enhancement:

### Runway Aleph 2

* **Credits:** 113
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** Runway
* **Avg. Duration:** \~514 seconds
* **Best for:** The latest Aleph for advanced style transfer and visual effects

### Kling Video O1 Edit Video

* **Credits:** 126
* **Rating:** ⭐⭐⭐ (3/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~222 seconds
* **Best for:** High-quality video editing and enhancement (Default model)

### Gemini Omni Flash 1.1 Edit Video

* **Credits:** 120
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** Instruction edits that keep the rest of the scene and stay consistent across successive edits

### FLUX 3 Edit Video

* **Credits:** 36
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** Fast natural-language edits that keep the clip's motion, timing and framing (input clips under 15 seconds)

### Gemini Omni Flash Edit Video

* **Credits:** 156
* **Rating:** ⭐⭐⭐⭐⭐ (5/5)
* **Provider:** fal.ai
* **Best for:** Conversational, iterative edits that preserve the rest of the scene

### LTX Video Retake

* **Credits:** 75
* **Rating:** ⭐⭐⭐ (3/5)
* **Provider:** fal.ai
* **Avg. Duration:** \~36 seconds
* **Best for:** Fast video retakes and modifications

## Supported Models

No additional supported models are currently available for this workflow.

## Tips for Best Results

1. **Start with Quality Source** — Better input video yields better enhancement results
2. **Define Effect Goals** — Specify what visual changes you want (lighting, style, color, etc.)
3. **Use Layer Controls** — Adjust intensity and frame ranges for subtle, story-driven effects
4. **Iterate Carefully** — VFX passes are non-destructive; you can revert to originals
5. **Match Project Aesthetic** — Ensure effects align with your project's visual style
6. **Preview Before Committing** — Review the enhanced result before saving

> 💡 **Tip:** VFX passes are non-destructive—you can always revert to the original render if the effect doesn't work as intended.

***

<CardGroup>
  <Card title="VFX (Visual Effects)" icon="sparkles" href="/tutorial/shot-studio/vfx-visual-effects">
    Learn how to apply visual effects to your shots
  </Card>

  <Card title="Shot Studio" icon="film" href="/tutorial/shot-studio/index">
    Complete guide to shot creation and enhancement
  </Card>
</CardGroup>


# Voice Models
Source: https://docs.automatstudio.com/models/voice-models

AI voice models available for text-to-audio generation

Automat Studio offers a diverse collection of AI voice models for generating dialogue, narration, and voiceovers. Each voice has unique characteristics, accents, and tonal qualities optimized for different creative projects.

<Note>
  These voice models are used for text-to-audio generation in Shot Studio and other audio workflows.
</Note>

## Available Voice Models

| Name | Sample | Description |
| - | - | - |
| Aria 🇺🇸 | [Listen](https://tkstwnhozchsymqhnkaw.supabase.co/storage/v1/object/public/media/voice-samples/voice_sample_aria.mp3) | A middle-aged female with an African-American accent. Calm with a hint of rasp. |
| Roger 🇺🇸 | [Listen](https://tkstwnhozchsymqhnkaw.supabase.co/storage/v1/object/public/media/voice-samples/voice_sample_roger.mp3) | Easy going and perfect for casual conversations. |
| Sarah 🇺🇸 | [Listen](https://tkstwnhozchsymqhnkaw.supabase.co/storage/v1/object/public/media/voice-samples/voice_sample_sarah.mp3) | Young adult woman with a confident and warm, mature quality and a reassuring, professional tone. |
| Laura 🇺🇸 | [Listen](https://tkstwnhozchsymqhnkaw.supabase.co/storage/v1/object/public/media/voice-samples/voice_sample_laura.mp3) | This young adult female voice delivers sunny enthusiasm with a quirky attitude. |
| Charlie 🇦🇺 | [Listen](https://tkstwnhozchsymqhnkaw.supabase.co/storage/v1/object/public/media/voice-samples/voice_sample_charlie.mp3) | A young Australian male with a confident and energetic voice. |
| George 🇺🇸 | [Listen](https://tkstwnhozchsymqhnkaw.supabase.co/storage/v1/object/public/media/voice-samples/voice_sample_george.mp3) | Warm resonance that instantly captivates listeners. |
| Callum 🇺🇸 | [Listen](https://tkstwnhozchsymqhnkaw.supabase.co/storage/v1/object/public/media/voice-samples/voice_sample_callum.mp3) | Deceptively gravelly, yet unsettling edge. |
| River 🇺🇸 | [Listen](https://tkstwnhozchsymqhnkaw.supabase.co/storage/v1/object/public/media/voice-samples/voice_sample_river.mp3) | A relaxed, neutral voice ready for narrations or conversational projects. |
| Liam 🇺🇸 | [Listen](https://tkstwnhozchsymqhnkaw.supabase.co/storage/v1/object/public/media/voice-samples/voice_sample_liam.mp3) | A young adult with energy and warmth - suitable for reels and shorts. |
| Charlotte 🇺🇸 | [Listen](https://tkstwnhozchsymqhnkaw.supabase.co/storage/v1/object/public/media/voice-samples/voice_sample_charlotte.mp3) | Sensual and raspy, she's ready to voice your temptress in video games. |
| Alice 🇬🇧 | [Listen](https://tkstwnhozchsymqhnkaw.supabase.co/storage/v1/object/public/media/voice-samples/voice_sample_alice.mp3) | Clear and engaging, friendly woman with a British accent suitable for e-learning. |
| Matilda 🇺🇸 | [Listen](https://tkstwnhozchsymqhnkaw.supabase.co/storage/v1/object/public/media/voice-samples/voice_sample_matilda.mp3) | A professional woman with a pleasing alto pitch. Suitable for many use cases. |
| Will 🇺🇸 | [Listen](https://tkstwnhozchsymqhnkaw.supabase.co/storage/v1/object/public/media/voice-samples/voice_sample_will.mp3) | Conversational and laid back. |
| Jessica 🇺🇸 | [Listen](https://tkstwnhozchsymqhnkaw.supabase.co/storage/v1/object/public/media/voice-samples/voice_sample_jessica.mp3) | Young and popular, this playful American female voice is perfect for trendy content. |
| Eric 🇺🇸 | [Listen](https://tkstwnhozchsymqhnkaw.supabase.co/storage/v1/object/public/media/voice-samples/voice_sample_eric.mp3) | A smooth tenor pitch from a man in his 40s - perfect for agentic use cases. |
| Chris 🇺🇸 | [Listen](https://tkstwnhozchsymqhnkaw.supabase.co/storage/v1/object/public/media/voice-samples/voice_sample_chris.mp3) | Natural and real, this down-to-earth voice is great across many use-cases. |
| Brian 🇺🇸 | [Listen](https://tkstwnhozchsymqhnkaw.supabase.co/storage/v1/object/public/media/voice-samples/voice_sample_brian.mp3) | Middle-aged man with a resonant and comforting tone. Great for narrations and advertisements. |
| Daniel 🇺🇸 | [Listen](https://tkstwnhozchsymqhnkaw.supabase.co/storage/v1/object/public/media/voice-samples/voice_sample_daniel.mp3) | A strong voice perfect for delivering a professional broadcast or news story. |
| Lily 🇬🇧 | [Listen](https://tkstwnhozchsymqhnkaw.supabase.co/storage/v1/object/public/media/voice-samples/voice_sample_lily.mp3) | Velvety British female voice delivers news and narrations with warmth and clarity. |
| Bill 🇺🇸 | [Listen](https://tkstwnhozchsymqhnkaw.supabase.co/storage/v1/object/public/media/voice-samples/voice_sample_bill.mp3) | Friendly and comforting voice ready to narrate your stories. |

***

## Choosing the Right Voice

When selecting a voice model for your project, consider:

* **Accent and Regionality** — Match the voice to your character's background or target audience
* **Tone and Emotion** — Choose voices that match the mood of your content
* **Age and Gender** — Select voices that fit your character profiles
* **Use Case** — Professional narrations, casual conversations, or character voices each benefit from different qualities
* **Sample Audio** — Always listen to samples before finalizing your choice

***

<CardGroup>
  <Card title="Text-to-Audio" icon="microphone" href="/models/text-to-audio">
    Learn about text-to-audio generation models
  </Card>

  <Card title="Shot Studio" icon="video" href="/tutorial/shot-studio">
    Use voice models in your shots
  </Card>
</CardGroup>


# Getting Help
Source: https://docs.automatstudio.com/support/contact

How to reach the Automat Studio team — in-app chat, email, or the contact form

There are three ways to reach us, and they all arrive in the same place. Pick whichever is easiest — you will not get a faster answer by using more than one.

***

## In the app

Click the floating **speech-bubble button** in the bottom-right corner of any page — or the speech-bubble icon in the left-hand menu — to open the support chat.

This is the best option if you already have an account. Because you are signed in, your plan and credit balance come through with the message, so you do not have to explain your setup before we can look at the problem.

## Email

Write to **[support@automatstudio.com](mailto:support@automatstudio.com)**.

Replies come from that same address, so you can keep the conversation in your inbox. If you reply to any email we send you — a welcome note, a receipt, a follow-up — it lands in the same place.

## Contact form

Use **Contact Us** in the website footer if you are not signed in, or do not have an account yet.

Give us an email address you actually read: it is the only way we can reply.

***

## What to include

You will get a useful answer faster if your first message covers:

* **What you were trying to do**, and what happened instead
* **Which project, scene or shot** it involves, by name
* **Which model** you were generating with, if it is a generation problem
* **When it happened**, roughly — it helps us find the right logs

Screenshots are welcome. If a generation failed, the shot name is usually enough for us to find it.

## When to expect a reply

We are a small team and we read every message ourselves. Most questions get an answer within two business days.

If your message is about **billing** — a charge you did not expect, a plan change that did not take effect, credits that seem wrong — say so in the first line. Those get looked at first, because they are usually time-sensitive.

***

## Try these first

Many questions have a written answer already, and it is faster than waiting for us:

<CardGroup>
  <Card title="Troubleshooting" icon="wrench" href="/support/troubleshooting">
    Fixes for the problems that come up most often
  </Card>

  <Card title="FAQ" icon="circle-question" href="/support/faq">
    How credits, plans, models and sharing work
  </Card>
</CardGroup>


# Frequently Asked Questions
Source: https://docs.automatstudio.com/support/faq

Common questions and answers about Automat Studio

## Getting Started

### What is Automat Studio?

Automat Studio is an AI filmmaking app for making your film. You can create characters, design locations, structure scenes, and generate shots—all within an integrated workflow that maintains visual consistency throughout your production. Watch, edit, and share your film with others.

### How do I get started?

Start by creating a new project, then follow the workflow: [Create a Project](tutorial/create-a-project) → [Create Characters](tutorial/character-studio) → [Create Locations](tutorial/location-studio) → [Create Scenes](tutorial/create-a-scene) → [Create Shots](tutorial/shot-studio). See our [complete guide](tutorial/index) for step-by-step instructions.

### Do I need to install anything?

No. Automat Studio runs entirely in your web browser — there's nothing to download and no GPU required, because all the generation happens on our servers. Any reasonably modern computer will do.

### Do I need to be good at writing AI prompts?

No. Automat Studio isn't a prompt box — it mirrors the way a real production works. You describe a character the way you'd write one in a screenplay, describe a location the way a production designer would, and pick the shot type, lens and camera movement from the same vocabulary a director uses on set.

***

## Features & Workflow

### What's the difference between scenes and shots?

**Scenes** are narrative units that combine a location with specific characters and represent a story moment. **Shots** are individual camera takes within a scene. You'll create multiple shots per scene to capture different angles, movements, and perspectives. Think of scenes as chapters and shots as paragraphs.

### Can I reuse characters and locations across different scenes?

Yes! Characters and locations are reusable assets. Once created, you can use them in multiple scenes and shots, which helps maintain visual consistency and saves time throughout your production. See [How to keep AI characters consistent across shots](/guides/consistent-ai-characters) for the full technique.

### Who owns the films and images I create?

You do. You retain full ownership of the AI-generated content you create on the platform, as set out in our [Terms of Service](/legal/terms) and [AI Content Policy](/legal/ai-content). We also do not use your content to train AI models, directly or indirectly.

### Can I build a project by describing it instead of filling in forms?

Yes. Every project has an assistant: open it from the sparkle button at the bottom right of any project page, or with **Create with AI** on the asset and scene pages, and describe what you want. It asks a question or two if it needs to, then creates the character, location, prop, wardrobe item, scene or shot in your project. It never spends credits on its own — images and video only generate after you confirm the price on a card in the chat. See the [Assistant](/tutorial/assistant) page.

### What are aesthetics and why should I use them?

Aesthetics define the visual style and color palette of your film. Applying an aesthetic to your project helps maintain a consistent look across all your shots. Automat Studio includes twelve default aesthetics, and Independent and Hollywood subscribers can create custom aesthetics using credits. Learn more in our [Aesthetics guide](tutorial/aesthetics).

***

## AI Models

### What's the difference between recommended and supported models?

**Recommended models** are the ones we suggest for the best experience in Automat Studio. Based on our testing, they produce the highest quality and most consistent results for narrative filmmaking workflows.

**Supported models** are available for use but may not provide the same level of quality or consistency as recommended ones. They can be useful for specific styles, experimentation, or particular creative purposes.

You can view all available models and their recommended uses in the [Models section](models/index).

***

## Credits & Subscriptions

### How do credits work?

Everything is paid for with a single **credits** balance. Each generation (character images, location images, first frames, videos, dialogue, lip sync, SFX, VFX, custom aesthetics) has a credit price that's deducted when it succeeds. Browsing, organizing projects, preparing assets, and failed or cancelled generations don't spend credits — you can cancel a running generation from its progress indicator for an immediate refund. Subscription plan credits refresh monthly; purchased credit packs never expire.

### What subscription plans are available?

Automat Studio offers three subscription tiers, each including monthly credits worth 20% more than the plan price:

* **Microbudget** (\$10/month, 1,200 credits) — For testing and smaller projects
* **Independent** (\$25/month, 3,000 credits) — For active projects with regular generation needs
* **Hollywood** (\$50/month, 6,000 credits) — For high-volume workflows and feature-length projects

See our complete [Credits & Subscriptions guide](/support/usage) for more details.

### How many projects can I have?

It depends on your plan: **free accounts have one active project**, Microbudget has 3, Independent has 10, and Hollywood is unlimited. Archived projects still count toward the limit, so on the free plan you'll need to either delete your existing project or [upgrade](https://automatstudio.com/subscriptions) before you can start a new one. See [Project Limits](/support/usage#project-limits) for details.

### Is there an unlimited plan?

There isn't currently an unlimited plan, but the higher subscription tiers are designed to support more exploratory and long-form workflows. If you find yourself consistently running out of credits, upgrade to a higher tier or buy a credit pack (purchased credits never expire).

### How do upgrades work?

Upgrades are prorated automatically. If you upgrade mid-cycle, you'll only be charged for the remaining time in your billing period and will receive additional plan credits for the rest of the month.

### How do downgrades work?

Downgrades are not prorated and take effect at the end of your current billing cycle. You'll retain full access to your current plan's features and credits until then, and your new plan will begin on your next billing date.

***

## Account

### I forgot my password — how do I reset it?

On the login screen, click **Forgot your password?**, enter the email address you signed up with, and we'll email you a link to set a new one. Open the link, choose a new password, and you'll be signed straight in.

Reset links are single-use and expire one hour after they're sent. If yours has expired or you've already used it, the page will offer to send a fresh one.

### I signed up with Google — where's my password?

There isn't one. Signing in with Google means Google handles your credentials, so there's nothing for us to reset. Just click **Sign in with Google** as usual; if you can't get into your Google account, recover it through Google.

***

## Feedback & Support

### How do I leave feedback or ask a question?

Click the floating *speech bubble button* in the bottom-right corner of any page (or the speech bubble icon on the left-hand menu) to open the support chat. You can ask a question or share feedback in the same place, and because you're signed in we can see your plan and credit balance without you having to explain your setup. We'd love to hear your thoughts—what works, what's missing, or what you'd love to see added. See [Getting Help](/support/contact) for the other ways to reach us.

### Where can I get help if I run into issues?

Check out our [Troubleshooting guide](/support/troubleshooting) for solutions to common problems. If you need more help, [Getting Help](/support/contact) covers every way to reach us — in-app chat, email, or the contact form on the website.


# Troubleshooting
Source: https://docs.automatstudio.com/support/troubleshooting

Solutions to common issues and problems in Automat Studio

If you're experiencing issues with Automat Studio, you're in the right place. This guide covers common problems and their solutions, organized by category.

***

## Getting Started Issues

### I can't sign in — I've forgotten my password

Click **Forgot your password?** on the login screen and enter the email address you signed up with. If that address has an account, we'll email a link that lets you set a new password.

**Solution:**

1. Click **Forgot your password?** below the Sign in button
2. Enter your email address and click **Send reset link**
3. Open the link in the email within one hour — it's single-use
4. Choose a new password; you'll be signed in automatically

<Note>
  **Signed up with Google?** You don't have an Automat Studio password — use **Sign in with Google** instead. Password reset only applies to email-and-password accounts.
</Note>

**If the link says it's expired:** reset links last one hour and can only be used once (some email apps consume the link by pre-fetching it). Request a new one from the same page and open it promptly. If the email hasn't arrived after a few minutes, check your spam folder.

### I can't create a scene without a location

This is expected behavior. You must create at least one location before you can create a scene, as scenes require a location to be assigned.

**Solution:**

1. Navigate to [Location Studio](tutorial/location-studio)
2. Create at least one location
3. Return to scene creation and select your location from the dropdown

<Note>
  **Prerequisite reminder:** Always create locations before scenes, and ensure you have characters before assigning them to scenes.
</Note>

### I can't create a custom aesthetic

Creating custom aesthetics requires an **Independent** or **Hollywood** plan, and each creation is paid for with [credits](/support/usage) — so the most common blockers are being on a lower plan or an empty credit balance.

**Solution:**

* Check your plan — upgrade to Independent or Hollywood to create custom aesthetics
* Check your credit balance in the app header or account settings
* Use one of the six default aesthetics available on all plans (they're free to apply)
* Top up with a credit pack or subscribe for monthly credits if you're running low

***

## Image Generation Problems

### Generated images don't match my description

AI image generation may not perfectly reflect every detail in your prompt, especially with shot attributes like lens type and camera angles. This is a current limitation that will improve over time.

**Solutions:**

1. **Be more specific** — Add more detail to your description about visual elements you want emphasized
2. **Try different models** — Click **Click here to change** below the generate button and experiment with different AI models (see [Models section](models/index) for recommendations)
3. **Generate multiple times** — Each generation is unique; try several iterations
4. **Refine iteratively** — Use previous generations as reference and refine your description based on what worked

> ⚠️ **Note:** Shot attributes (like lens and shot type) may not be precisely reflected in generated images due to current AI limitations—but this will improve over time!

### Character or location images look inconsistent

If your characters or locations vary too much between generations, here are some strategies:

**Solutions:**

1. **Use more detailed descriptions** — Include specific physical attributes, clothing, colors, and distinctive features
2. **Save and reuse** — Once you get a character or location you like, save it and reuse it across scenes
3. **Apply project aesthetics** — Using a consistent aesthetic can help maintain visual coherence (see [Aesthetics guide](tutorial/aesthetics))
4. **Reference previous generations** — Some models work better when you reference or build upon previous successful generations

### The failure notice says my prompt was blocked by the model's content-safety filter

Some AI models run their own safety checker on top of Automat Studio's content filter, and reject prompts they consider unsafe — including prompts that name trademarked characters or real people. When that happens, the failure notice tells you the image model's content-safety filter blocked the prompt, and no credits are spent (see [Usage & Credits](/support/usage)).

**Solutions:**

1. **Reword the prompt** — describe the look you want in your own words rather than naming a copyrighted character or a real person; our [Terms of Service](/legal/terms) ask you to design your own
2. **Switch models** — click **Click here to change** below the generate button; each model applies its own rules, so another model may accept the same description
3. **Simplify** — remove the part of the description the model is likely objecting to and add detail back gradually

### Image generation is slow or stuck

Image generation typically takes 30 seconds to 2 minutes, depending on the model and complexity.

**Solutions:**

1. **Be patient** — Wait at least 2-3 minutes before assuming it's stuck
2. **Check your connection** — Ensure you have a stable internet connection
3. **Try a different model** — Some models are faster than others
4. **Clear and retry** — Cancel the generation and try again with a simpler prompt
5. **Check server status** — High server load can cause delays

***

## Video Generation Issues

### Video generation takes too long

Video generation typically takes 2-3 minutes. You can navigate away from the page while it processes in the background.

**Solutions:**

1. **Wait patiently** — Video generation is more resource-intensive than image generation
2. **Work on other shots** — Automat processes in the background, so you can continue working
3. **Check the progress indicator** — Look for status updates in the shot interface
4. **Reduce shot duration** — Shorter clips generate faster

<Note>
  You can navigate away from the page while rendering—Automat will keep processing in the background. Check back later to see if your video is ready.
</Note>

### Generated video doesn't animate smoothly

Motion quality can vary based on the source image and the model used.

**Solutions:**

1. **Start with a better first frame** — Ensure your first frame is clear and well-composed
2. **Adjust motion settings** — Experiment with motion strength and camera path settings
3. **Try different models** — Some models handle motion better than others
4. **Regenerate** — Each generation is different; try multiple attempts
5. **Use Generate Video Direct** — This alternative method might work better for your specific shot (see [Generate Video Direct](tutorial/shot-studio/generate-video-direct))

### Video won't generate from my first frame

**Solutions:**

1. **Ensure first frame is saved** — Make sure you've clicked Save on your first frame
2. **Check frame selection** — Confirm the correct frame is selected in the shot gallery
3. **Verify shot details** — Ensure duration, camera movement, and other settings are configured
4. **Try regenerating the first frame** — Sometimes a different still works better for animation

***

## Aesthetic Problems

### Custom aesthetic processing failed

If your custom aesthetic creation fails, try these steps:

**Solutions:**

1. **Check image quality** — Ensure your uploaded images are clear and in supported formats (JPG, PNG, WebP, GIF)
2. **Reduce image count** — Try processing with fewer images (start with 2-3 instead of 5)
3. **Use consistent images** — Images should share similar visual styles for best results
4. **Check file size** — Very large images may cause processing issues; resize if necessary
5. **Retry the process** — Sometimes a temporary server issue; try again after a few minutes

> 💡 **Tip:** The more consistent your reference images are in style, the more cohesive your resulting aesthetic will be. Choose images that share similar visual qualities.

### Aesthetic isn't applying to my project

**Solutions:**

1. **Check project settings** — Ensure the aesthetic is selected in your project settings
2. **Re-apply aesthetic** — Try changing to "None" and then back to your aesthetic
3. **Verify it's processed** — Custom aesthetics must be fully processed before use
4. **Create new project** — Some aesthetics work better when selected during project creation

***

## Model Selection Issues

### I don't know which model to use

**Solutions:**

1. **Start with recommended models** — These are tested for best results in Automat Studio (see [Models guide](models/index))
2. **Use supported models for experimentation** — Try different models to find what works for your style
3. **Look below the generate button** — the model in use is shown there; click **Click here to change** to see and select from available models
4. **Check model documentation** — Each model has different strengths; see the [Models section](models/index) for details

<Note>
  **Recommended models** produce the best results for narrative filmmaking. **Supported models** may not provide the same level of quality but can be useful for specific styles or experimentation.
</Note>

### Model keeps producing poor results

**Solutions:**

1. **Switch to recommended models** — Recommended models are optimized for Automat Studio workflows
2. **Refine your prompts** — Better descriptions often yield better results regardless of model
3. **Try multiple models** — Different models excel at different things; experiment to find the best fit
4. **Check model status** — Some models may have temporary issues; try again later

***

## File Upload & Data Issues

### Can't upload images for custom aesthetics

**Solutions:**

1. **Check file format** — Supported formats are JPG, PNG, WebP, and GIF. Other types (including HEIC from an iPhone and SVG) are refused by the file picker — convert them first
2. **Verify file size** — Very large files may fail; try resizing images to under 10MB each
3. **Check upload limit** — Maximum of 5 images for custom aesthetics
4. **Try different browser** — Some browsers handle uploads better than others
5. **Check your plan** — Ensure you're on a plan that supports custom aesthetics

### Can't upload a reference image for an asset

**Solutions:**

1. **Check file format** — Supported formats are JPG, PNG, and WebP
2. **Verify file size** — Reference images are limited to 10MB each; resize larger files
3. **Upload one image** — Character, location, prop, and wardrobe forms take a single reference image
4. **Try different browser** — Some browsers handle uploads better than others

### Lost work or changes not saving

**Solutions:**

1. **Save frequently** — Always click Save after making changes to characters, locations, scenes, or shots
2. **Check internet connection** — Unstable connections can prevent saves
3. **Use shot history** — Previous renders are stored; you can revert to earlier versions
4. **Don't navigate away too quickly** — Wait a moment after clicking Save before leaving

***

## Workflow & Navigation

### I'm confused about the workflow order

The recommended workflow is:

1. [Create a Project](tutorial/create-a-project)
2. [Create Characters](tutorial/character-studio)
3. [Create Locations](tutorial/location-studio)
4. [Create Scenes](tutorial/create-a-scene)
5. [Create Shots](tutorial/shot-studio)

**Within each shot:**

* Generate First Frame
* Generate Video
* Add Dialogue and Lipsync
* Apply VFX
* Add SFX

See the complete [workflow guide](tutorial/index) for detailed instructions.

### Can't find where to create a shot

**Solutions:**

1. **Start with a scene** — Shots are created within scenes
2. **Navigate to scene view** — Open your scene and look for "Create Shot" button
3. **Check you have prerequisites** — Ensure your scene has a location and characters assigned
4. **See [Create a Scene](tutorial/create-a-scene)** — Step 4 covers shot creation

***

## Performance & Technical Issues

### App is slow or unresponsive

**Solutions:**

1. **Refresh the page** — Sometimes a simple refresh resolves temporary issues
2. **Check browser** — Ensure you're using a modern, up-to-date browser
3. **Clear cache** — Clear your browser cache and cookies
4. **Close other tabs** — Reduce system load by closing unnecessary browser tabs
5. **Check internet speed** — Slow connections can make the app feel sluggish
6. **Try different browser** — Some browsers perform better than others

### Images or videos won't load

**Solutions:**

1. **Refresh the page** — Often resolves loading issues
2. **Check internet connection** — Ensure you have a stable connection
3. **Wait a moment** — Large files may take time to load
4. **Try different browser** — Browser compatibility issues can prevent media loading
5. **Clear browser cache** — Corrupted cache can prevent media from loading

***

## Still Need Help?

If you've tried these solutions and still need assistance:

<CardGroup>
  <Card title="Frequently Asked Questions" icon="circle-question" href="/support/faq">
    Browse common questions and answers about Automat Studio
  </Card>

  <Card title="Getting Help" icon="envelope" href="/support/contact">
    In-app chat, email, or the contact form — and what to include so we can
    help on the first reply
  </Card>

  <Card title="Complete Workflow Guide" icon="book" href="/tutorial/index">
    Review the full step-by-step guide for creating your film
  </Card>
</CardGroup>


# Credits & Subscriptions
Source: https://docs.automatstudio.com/support/usage

Understanding how credits work and what each subscription plan includes

Automat Studio uses a single **credits** balance to pay for generation. Every model has a credit price, every generation spends credits, and subscriptions top your credits up every month. This guide explains how credits work and what each plan offers.

***

## How Credits Work

Usage is based on **generation activity**, not time spent in the app. Each generation has a credit price (shown on the generate button and in the model picker) that is deducted from your credit balance.

Your balance can come from three places:

* **Free signup credits** — every new account starts with 500 credits.
* **Plan credits** — your subscription includes a set number of credits every month (see the plans table below). Plan credits **refresh every month** and do not carry over. This is true on yearly billing too: yearly plans receive the same credit refresh at each monthly anniversary of the subscription.
* **Purchased credits** — one-off [credit packs](https://automatstudio.com/credits) you can buy at any time, with or without a subscription. Purchased credits **never expire**.

The app shows one combined credit counter. Plan (and free signup) credits are spent first; purchased credits are used once those run out. Your account page shows the breakdown of plan vs purchased credits.

<Note>
  Credits do not unlock features from higher subscription tiers. Feature access is determined by your active subscription plan, while credits pay for generation usage.
</Note>

### What Counts as Usage

The following activities spend credits:

* **Character images** — Generating new character portraits
* **Location images** — Generating new location visuals
* **First frames** — Generating still images for shots
* **Videos** — Generating video clips from first frames
* **Dialogue & lip sync** — Generating character voice audio and syncing it to video
* **SFX & VFX** — Generating sound effects and visual effects for your shots
* **Custom aesthetics** — Creating your own visual styles from reference images

### What Doesn't Count as Usage

The following activities are **free** and don't spend credits:

* **Browsing** — Viewing your projects, characters, locations, and shots
* **Organizing** — Creating projects, scenes, and structuring your workflow
* **Preparing assets** — Writing descriptions, adjusting settings, and planning shots
* **Failed generations** — If a generation fails (including content filter failures), no credits are spent. When the AI model itself rejects a prompt on safety grounds, the failure notice says so and suggests rewording the prompt or switching to a different model
* **Cancelled generations** — If you cancel a generation while it's running (the Cancel button under the progress indicator), the credits are refunded immediately
* **Timed-out generations** — Every generation has a time limit (5 minutes for images and audio, 20 minutes for video). If the model hasn't answered by then, the generation is stopped, marked as timed out, and its credits are returned automatically
* **Support refunds** — If support returns credits to you (for example after a generation came out clearly wrong), they go straight back into your balance and appear on your account's Transactions page as a **Refund**

<Note>
  Only successful generations spend credits. You won't be charged for failed attempts, cancelled generations, or system errors.
</Note>

***

## Subscription Plans

Each subscription includes monthly credits worth **20% more than the plan's price**, plus access to plan-gated features. Higher tiers are designed to support more exploratory and long-form workflows. Every plan is available with monthly or yearly billing (yearly saves 20%) — credits refresh monthly on both.

| Plan | Price | Monthly Credits | Active Projects | Best For |
| - | - | - | - | - |
| **Microbudget** | \$10/month | 1,200 | 3 | Testing the platform and smaller projects |
| **Independent** | \$25/month | 3,000 | 10 | Active projects with regular generation needs |
| **Hollywood** | \$50/month | 6,000 | Unlimited | High-volume workflows and feature-length projects |

<Note>
  There is currently no unlimited plan. If you find yourself consistently running out of credits, upgrade to a higher tier or top up with a credit pack.
</Note>

<Warning>
  Unused plan credits do not carry over when the monthly refresh lands, and expire if you cancel your subscription. Purchased credits remain available until used.
</Warning>

### Project Limits

Each plan caps how many projects you can have at once. **Free accounts have one active project**; Microbudget raises that to 3, Independent to 10, and Hollywood has no limit. When you're at your limit, the **New Project** button on your dashboard is disabled and explains the limit when you hover over it.

**Archived projects still count toward the limit.** Archiving tidies a project away from your dashboard but keeps its slot; deleting a project frees it. So to start something new on the free plan, either delete the project you've finished with or [upgrade](https://automatstudio.com/subscriptions).

***

## Upgrading Your Plan

You can upgrade your subscription at any time. Upgrades are **prorated automatically**—you’ll only be charged the difference for the remaining time in your billing cycle.

Your upgraded plan takes effect immediately, and additional plan credits are added for the rest of the month. In some cases, this may be a partial credit increase depending on when you upgrade.

**To upgrade:**

1. Visit your account settings or the [subscriptions page](https://automatstudio.com/subscriptions)
2. Select your desired plan
3. Confirm the upgrade

<Note>
  To avoid double-charging, both billing and plan credits may be prorated when upgrading mid-cycle.
</Note>

### Example

If you start on the $10 plan and upgrade to the $25 plan partway through the month, you’ll only be charged the difference for the remaining time. Your plan credits for that month will increase accordingly.

At the start of your next billing cycle, your full 3,000 monthly credits will be applied.

***

## Downgrading Your Plan

Downgrades are **not prorated** and take effect at the end of your current billing cycle. You'll retain full access to your current plan's features and credits until then.

**To downgrade:**

1. Visit your account settings or the [subscriptions page](https://automatstudio.com/subscriptions)
2. Select a lower-tier plan
3. Confirm the downgrade

Your new plan and monthly credits will begin on your next billing date. While a downgrade is scheduled, your account and subscriptions pages show the upcoming change and its effective date — no need to submit it again.

***

## Tips for Managing Credits

* **Plan before you generate** — Write detailed descriptions and prepare your assets before generating to get better results on the first try
* **Use the right model** — Choosing appropriate models (see [Models guide](/models/index)) can help you achieve desired results with fewer regenerations
* **Save successful generations** — Once you create a character or location you like, save it and reuse it across scenes
* **Apply aesthetics** — Using consistent [aesthetics](/tutorial/aesthetics) helps maintain visual coherence, reducing the need for regeneration
* **Top up strategically** — Purchased credits never expire, so a credit pack is a safe way to cover overflow beyond your plan's monthly credits

***

## Questions?

<CardGroup>
  <Card title="Frequently Asked Questions" icon="circle-question" href="/support/faq">
    Browse common questions about Automat Studio features
  </Card>

  <Card title="Getting Help" icon="envelope" href="/support/contact">
    Reach out if you have questions about your plan or credits
  </Card>
</CardGroup>


# Create Aesthetics
Source: https://docs.automatstudio.com/tutorial/aesthetics

Create custom visual styles and palettes for your film projects

Aesthetics define the visual style and color palette of your film. By applying an aesthetic to your project, you can maintain a consistent look and feel across all your shots, characters, and locations. Automat Studio analyzes reference images to create custom aesthetics that match your artistic vision.

<Note>
  Creating custom aesthetics requires an **Independent** or **Hollywood** plan, and each creation is paid for with [credits](/support/usage). All plans have access to the default aesthetics listed below.
</Note>

## Default Aesthetics

All Automat Studio plans include access to these twelve predefined aesthetics:

* **Pastel Americana** — Soft, nostalgic tones inspired by mid-century American design
* **Hand-Drawn** — Artistic, illustrated look with sketch-like qualities
* **90s European Art House** — Sophisticated, cinematic style reminiscent of European film movements
* **Digital Animation** — Clean, vibrant aesthetic perfect for animated content
* **Dark Academia** — Moody, intellectual atmosphere with rich, subdued colors
* **Neon Cyberpunk** — Futuristic, high-contrast palette with electric colors and urban edge
* **Painterly Neon** — Kinetic mixed-media comic-book animation with brushstrokes, halftones and neon pop
* **Watercolor Anime** — Hand-painted anime warmth with lush watercolor backgrounds and cel-shaded characters
* **Film Noir** — Classic 1940s black-and-white cinema with hard shadows and rain-slicked streets
* **Storybook Symmetry** — Perfectly symmetrical, pastel-coordinated whimsy with deadpan storybook charm
* **Stop-Motion** — Handcrafted puppet-and-miniature look with tactile clay, felt and tilt-shift depth
* **Retro VHS** — Nostalgic 1980s camcorder footage with analog scanlines, chroma bleed and halation

You can select any of these defaults when creating a new project, or choose "None" to build your style organically.

***

## Creating a Custom Aesthetic

Custom aesthetics allow you to train Automat Studio on your own visual references, creating a unique style that can be applied across your entire project.

### Step 1: Access Aesthetics

* Click on the **Aesthetics** icon in the main navigation menu

### Step 2: Create New Aesthetic

* Click **"Create New Aesthetic"**
* Upload up to 5 reference images (drag and drop or click to upload)
  * Choose images that represent the visual style, color palette, and mood you want to achieve
  * These can be screenshots, stills from films, photographs, or artwork

### Step 3: Process Images

* Click **"Process Images"** to analyze your references
* Automat Studio's AI will analyze the images and create a custom aesthetic based on:
  * Color palettes and tones
  * Lighting styles
  * Composition patterns
  * Visual mood and atmosphere

### Step 4: Use Your Aesthetic

* Once processed, your new aesthetic will be available in the project creation dropdown
* Select it when creating a new project, or apply it to existing projects through project settings

> 💡 **Tip:** The more consistent your reference images are in style, the more cohesive your resulting aesthetic will be. Choose images that share similar visual qualities for best results.

***

## Applying Aesthetics to Projects

You can set an aesthetic when creating a new project (see [Create a Project](tutorial/create-a-project)), or change it later in your project settings. The aesthetic will influence all generated images and videos within that project, helping maintain visual consistency.

***

<CardGroup>
  <Card title="Create a Scene" icon="arrow-left" href="/tutorial/create-a-scene">
    Previous: create scenes
  </Card>

  <Card title="Shot Studio" icon="video" href="/tutorial/shot-studio/index">
    Next: create shots
  </Card>
</CardGroup>


# Assistant
Source: https://docs.automatstudio.com/tutorial/assistant

Build your project by describing it: the assistant creates characters, locations, props, wardrobe, scenes and shots for you

The assistant is a chat panel inside every project. Instead of filling in a form, you describe what you want — "a 24-year-old nurse called Mara, blond, blue eyes, tired but sharp" — and the assistant asks a question or two if it needs to, then creates it in your project. You can carry on in the studios as usual; everything it makes is an ordinary character, location, prop, wardrobe item, scene or shot.

## Opening it

There are two ways in:

* The **sparkle button** at the bottom right of any project page, next to the Help button. On a phone it fills the screen; on a desktop it sits beside your work.
* **Create with AI** at the top of the Characters, Locations, Props, Wardrobe and Scenes pages, and **Ask the assistant** at the bottom of each studio. These open the panel already knowing where you are, so "make her older" inside a character's studio means that character.

## What it can do

* **Create and edit assets.** Characters (name, age, description, voice), locations (interior or exterior, time of day), props and wardrobe. It writes descriptions the way a screenplay introduces things, which is what the image models respond to best — and you can edit anything afterwards in the studio.
* **Set up scenes and shots.** A scene needs a location, so if the one you describe does not exist yet it will offer to create it first. Shots get a first-frame description with the right `@mentions` for the characters, props and wardrobe in frame, a video description, and dialogue with its speaker. See [Mention Characters, Props & Wardrobe](/tutorial/shot-studio/mention-assets) for why mentions matter.
* **Offer to generate images and video.** After creating something it will quote the price and put a card in the chat.

## It never spends credits on its own

Images and video cost credits, and the assistant cannot start a generation itself. When it offers one, a card appears in the chat showing the model, the price and your balance, with a **Generate for N credits** button. Nothing happens until you press it. Pressing it does exactly what the studio's generate button does, so the same rules and prices apply. If you would rather not, choose **Not now**.

<Note>
  Creating and editing rows — characters, scenes, shots and so on — is free. Only generations cost credits, and only after you confirm.
</Note>

## What it cannot do

* Delete, archive or hide anything. Use the project pages for that.
* Reorder scenes or shots — drag and drop them on the Scenes page.
* Change your plan, credits or account settings. Those are on the Credits and Account pages.

If you ask for one of these it will tell you where to go.

## Good things to know

* **It asks at most a question or two.** If you give it enough to go on, it will not ask anything. If it fills in a detail you did not want, just tell it, or edit the asset in its studio.
* **The chat is not saved.** It lasts while you are in the project; leaving the project or reloading the page starts fresh. What it created in your project stays.
* **Names must be unique within a project**, because shots refer to assets by name. If you ask for a second "Mara" it will ask whether you meant to update the first.
* **The content filter still applies.** If a description is blocked, the assistant will say so in the filter's words and ask for a different wording.
* **Stop** interrupts a reply mid-stream; anything already created stays.

***

<CardGroup>
  <Card title="Character Studio" icon="user" href="/tutorial/character-studio">
    Edit what the assistant created
  </Card>

  <Card title="Shot Studio" icon="video" href="/tutorial/shot-studio/index">
    Generate the shots it set up
  </Card>
</CardGroup>


# Character Studio
Source: https://docs.automatstudio.com/tutorial/character-studio

Create and manage characters for your film

## Quick Start Video

<iframe title="Automat Studio Character Studio" />

***

## 2. Create a Character

Characters are the people who appear in your film. Each character can be reused across multiple scenes and shots, ensuring visual consistency throughout your production.

<img alt="Character Studio overview" />

The studio opens as a dialog with the controls on the left and the image on the right. Prefer the picture big? **Switch to fullscreen view** (bottom left, next to Watch Tutorial) opens the same character with the image filling the screen and the controls floating on top — see [Fullscreen view](#fullscreen-view) below. Whatever you have typed comes with you.

### Step 1: Enter Character Details

* On the Characters page, click **Create New Character**
* Enter the character's **Name**
* Enter the character's **Age**
* Write a detailed **Description** — include physical appearance (hair, height, build, distinctive features), clothing style, personality traits, and any other visual characteristics that will help the AI generate an accurate representation
* **Select a voice** — You can select the voice your character will have here. Click **Play Sample** to hear a sample. This voice will be used when generating dialogue

Your details also save on their own a few seconds after you stop typing (**Saving…**, then **Saved**, at the top of the dialog).

<Note>
  **Building a character from a photo?** Switch the New / Edit / **Reference Image** control to Reference Image and upload it there — that is the only place a photo reaches the model. Describing an upload in the Description field does nothing, because on the New method there is no image to reference. Note that the Reference Image method uses your photo as a *likeness reference* and generates a new studio portrait from it, rather than reproducing the photo itself — see [Reference-to-Image](/models/reference-to-image).
</Note>

### Step 2: Choose Draft or Final quality

Under the generate button you'll find a **Draft / Final** quality switch:

* **Draft** — fast and inexpensive, ideal while you're iterating on the description
* **Final** — the best image quality on offer, for when you're happy with the direction

The generate button always shows exactly what a generation will cost with the current choice. The model line below the switch shows which AI model is in use — click **Click here to change** to pick any supported model instead; the switch clears until you tap a quality again, and your quality choice carries over when you swap between the New / Edit / Reference Image methods.

### Step 3: Generate Character Image

* Click **Generate Character Image** — the details you entered will guide the AI image generation
* Each character is generated as a **reference sheet** — three consistent views of your character (a front portrait, a side profile, and a full-body shot) in a single image
* Review the generated image and refine your description if needed
* Generate again until you're satisfied with the result
* Click **Save and Close** to add the character to your project

<Note>
  Automat uses your character's reference sheet when generating scenes and shots. The extra angles help keep your character looking consistent across different shot types and camera angles — including profile and wide shots, where a single front-on portrait tends to drift.
</Note>

### Editing an existing image

Switch the New / Edit / **Reference Image** control to **Edit** to change the active image rather than start over: describe the change ("give her a red scarf", "rain-soaked", "older") and click **Edit Character Image**. The active image is the source, so set the generation you want to build on as active first.

### Managing Generations

Once you've generated character images, you can manage them with the following features:

* **Navigate through generations** — Use the back and next buttons under the image to browse through all generated versions
* **Provide feedback** — Click the thumbs up and thumbs down buttons to give feedback on generated images
* **Active image** — The active image will be used when creating shots and scenes. New generations are automatically set as active
* **Set active image** — To set a different image as active, navigate to it using the back/next buttons and click **Set Active** underneath the image
* **Hide a generation** — Click the eye button to tuck away a generation you no longer want to see, without losing it. To bring it back, click the show-hidden (eye) button next to the navigation counter, navigate to the hidden generation, and click **Unhide**
* **Delete a generation** — Click the trash button to remove a generation permanently. If you might want it back later, hide it instead

### Fullscreen view

Click **Switch to fullscreen view** at the bottom of the dialog to work with the image filling the whole screen. Everything above works the same way; the layout is different:

* **Controls panel** — Name, the New / Edit / Reference control, the description, Draft / Final and the generate button float in a panel at the left — drag it by the grip along its top edge to uncover the picture, or drag it to the left or right side of the window to dock it there as a rail (a guide appears when it is close enough; the picture switches to Fit and shrinks to make room, and dragging it away floats it again — double-click the grip to put it back). **More — age & voice** holds the Age and Voice fields. On a phone the panel is a sheet that slides up from the bottom — tap the sliders icon in the top bar to open it
* **Takes strip** — every generation runs along the bottom as a strip of thumbnails. Click one, or use the **←** and **→** keys, to put it on screen; the take with the green tick is the active image
* **Action bar** — thumbs up/down, favourite, info, hide, delete and **Set Active** sit in a bar just above the strip. The heart and eye buttons at the left of the strip filter to favourites or reveal hidden takes
* **Fill or Fit** — the top-right switch crops the image to fill the screen or fits the whole image inside it. Click the image itself to open it full size
* **Closing** — the **×** in the top right (or **Esc**) saves anything unsaved and returns you to your characters. **Switch to classic view** at the bottom of the panel goes back to the dialog

<Note>
  **Tip:** The AI model in use is shown just below the generate button — click **Click here to change** to experiment with different models for varied visual styles and quality levels.
</Note>

***

<CardGroup>
  <Card title="Create a Project" icon="arrow-left" href="/tutorial/create-a-project">
    Previous: project setup
  </Card>

  <Card title="Location Studio" icon="location-dot" href="/tutorial/location-studio">
    Next: create your location
  </Card>
</CardGroup>


# Create a Project
Source: https://docs.automatstudio.com/tutorial/create-a-project

Start a new film project in Automat Studio

A project is your workspace for a complete film. It organizes all your characters, locations, scenes, and shots in one place, making it easy to manage and collaborate on your production.

## 1. Create a Project

From the home page or dashboard, click **Create New Project** and fill in the following details:

* **Name** — Choose a descriptive name for your project (e.g., "Sci-Fi Short: The Visitor" or "Documentary: City Life")
* **Synopsis** — Write a brief overview of your film's story, theme, or concept. This helps set the tone and can guide your creative decisions throughout production
* **Aesthetic** — Select a predefined visual style (if available) that matches your vision, or choose "None" to build your style organically as you create shots

Click **Create** to initialize your project workspace.

<Note>
  You can always edit the project name and synopsis later, but setting a clear aesthetic early can help maintain visual consistency across all your shots.
</Note>

<Note>
  **Free accounts have one active project** (Microbudget 3, Independent 10, Hollywood unlimited). Archiving a project does not free its slot — only deleting it does — so on the free plan, delete your finished project or [upgrade](https://automatstudio.com/subscriptions) before starting a new one. See [Project Limits](/support/usage#project-limits).
</Note>

***

<CardGroup>
  <Card title="Tutorial Overview" icon="arrow-left" href="/tutorial/index">
    Back to tutorial overview
  </Card>

  <Card title="Character Studio" icon="users" href="/tutorial/character-studio">
    Next: create your characters
  </Card>
</CardGroup>


# Create a Scene
Source: https://docs.automatstudio.com/tutorial/create-a-scene

Combine locations, characters, and props into narrative scenes

Scenes are the building blocks of your film's narrative structure. Each scene combines a location with characters and props, representing a story moment or sequence. You'll create multiple shots within each scene to capture different angles and perspectives.

## 5. Create a Scene

> ⚠️ **Prerequisite:** You must create at least one location before you can create a scene.

### Step 1: Enter Scene Details

* From the Scenes page, click **Create New Scene**
* Select a **Location** from the dropdown menu of all locations you've created in your project
* Select **Int** (Interior) or **Ext** (Exterior) to match your location type
* Select the **Time of Day** (Day or Night) for this scene
* (Optional) Add a **Synopsis** describing what happens in this scene

> 💡 **Tip:** Use the location you just created and add the characters who will appear together in this story moment.

### Step 2: Save the Scene

* Click **Save** to create the scene
* Your selected location will appear on the scene page

### Step 3: Add Characters

* From the multi-select character list, choose which characters appear in this scene
* Click **Save** after making your character selections

### Step 4: Add Props

* From the multi-select prop list, choose which props appear in this scene
* Click **Save** after making your prop selections

<Tip>
  Prefer to describe it? The [Assistant](/tutorial/assistant) can set up a scene — and the location it needs — from a sentence, then add shots to it.
</Tip>

### Step 5: Create Shots

Once your scene is set up with a location and characters, you're ready to create individual shots:

* From the view scene page, click **"Create Shot"**
* This will add a new shot to your scene and open the Shot Studio
* See the [Shot Studio guide](/tutorial/shot-studio/index) for complete details on working with shots

***

<CardGroup>
  <Card title="Wardrobe Studio" icon="arrow-left" href="/tutorial/wardrobe-studio">
    Previous: create costumes and outfits
  </Card>

  <Card title="Aesthetics" icon="palette" href="/tutorial/aesthetics">
    Next: explore visual styles
  </Card>
</CardGroup>


# Introduction
Source: https://docs.automatstudio.com/tutorial/index

Complete guide to creating your film in Automat Studio

Welcome to Automat Studio's filmmaking workflow! This guide will walk you through the entire process of bringing your vision to life, from initial project setup to final rendered shots.

## Quick Start Video

**How to Visualize Your Film With Automat Studio (In 90 Seconds)**

<iframe title="How to Visualize Your Film With Automat Studio (In 90 Seconds)" />

***

## The Automat Studio Workflow

Follow these steps in order to create your film:

### 1. [Create a Project](tutorial/create-a-project)

Set up your workspace with a name, synopsis, and aesthetic style.

### 2. [Create Characters](tutorial/character-studio)

Define the people who will appear in your film with detailed descriptions and generated images.

### 3. [Create Locations](tutorial/location-studio)

Build the environments where your scenes take place, specifying interior/exterior and time of day.

### 4. [Create Props](tutorial/prop-studio)

Add objects and items that appear in your scenes with detailed descriptions and generated images.

### 5. [Create Wardrobe](tutorial/wardrobe-studio)

Design the costumes and outfits your characters wear, with an optional link to group them by character.

### 6. [Create Scenes](tutorial/create-a-scene)

Combine locations, characters, and props into scenes that structure your narrative.

### 7. [Create Shots](tutorial/shot-studio)

Break down scenes into individual shots, each with specific camera angles and movements.

Within each shot, you'll work through these steps:

* **[Generate First Frame](tutorial/shot-studio/generate-first-frame)** — Create the fist frame of your shot
* **[Generate Video](tutorial/shot-studio/generate-video)** — Animate the first frame into motion
* **[Dialogue and Lipsync](tutorial/shot-studio/dialogue-lipsync)** — Add voice performances with synchronized lip movement
* **[VFX (Visual Effects)](tutorial/shot-studio/vfx-visual-effects)** — Layer cinematic effects and enhancements
* **[SFX (Sound Effects)](tutorial/shot-studio/sfx-sound-effects)** — Design soundscapes that match the action

Repeat this process for each shot until your film is complete!

### 7. [Share and Export](tutorial/share-and-export)

Share your completed project with others via shareable links, or export your media assets for use in external editing software.

***

<CardGroup>
  <Card title="Create a Project" icon="folder-open" href="tutorial/create-a-project">
    Start by creating your project workspace
  </Card>
</CardGroup>


# Location Studio
Source: https://docs.automatstudio.com/tutorial/location-studio

Create and manage locations for your film

## Quick Start Video

<iframe title="Automat Studio Location Studio" />

***

## 3. Create a Location

Locations define where your scenes take place. Each location can be reused across multiple scenes, saving time and maintaining visual consistency throughout your film.

Prefer the picture big? **Switch to fullscreen view** (bottom left of the dialog, next to Watch Tutorial) opens the same location with the image filling the screen and the controls floating on top — see [Fullscreen view](#fullscreen-view) below. Whatever you have typed comes with you.

<img alt="Location Studio overview" />

### Step 1: Enter Location Details

* Click **Create New Location**
* Enter a descriptive **Name** (e.g., "Downtown Coffee Shop" or "Mountain Cabin")
* Select **Int.** (Interior) or **Ext.** (Exterior) to indicate the location type
* Select **Day** or **Night** for the lighting condition
* Write a detailed physical **Description** of the location — include architectural details, atmosphere, mood, and key visual elements

### Step 2: Choose Draft or Final quality

Under the generate button you'll find a **Draft / Final** quality switch:

* **Draft** — fast and inexpensive, ideal while you're iterating on the description
* **Final** — the best image quality on offer, for when you're happy with the direction

The generate button always shows exactly what a generation will cost with the current choice. The model line below the switch shows which AI model is in use — click **Click here to change** to pick any supported model instead; the switch clears until you tap a quality again, and your quality choice carries over when you swap between generation methods.

### Step 3: Generate Location Image

* Click **Generate Location Image** — your description will guide the AI image generation
* Review the generated image and refine your description if needed
* Generate again until you're satisfied with the result
* Click **Save and Close** to add the location to your project

### Managing Generations

Once you've generated location images, you can manage them with the following features:

* **Navigate through generations** — Use the back and next buttons under the image to browse through all generated versions
* **Provide feedback** — Click the thumbs up and thumbs down buttons to give feedback on generated images
* **Active image** — The active image will be used when creating shots and scenes. New generations are automatically set as active
* **Set active image** — To set a different image as active, navigate to it using the back/next buttons and click **Set Active** underneath the image
* **Hide a generation** — Click the eye button to tuck away a generation you no longer want to see, without losing it. To bring it back, click the show-hidden (eye) button next to the navigation counter, navigate to the hidden generation, and click **Unhide**
* **Delete a generation** — Click the trash button to remove a generation permanently. If you might want it back later, hide it instead

### Fullscreen view

Click **Switch to fullscreen view** at the bottom of the dialog to work with the image filling the whole screen. Everything above works the same way; the layout is different:

* **Controls panel** — Name, the New / Edit / Reference control, the description, Draft / Final and the generate button float in a panel at the left — drag it by the grip along its top edge to uncover the picture, or drag it to the left or right side of the window to dock it there as a rail (a guide appears when it is close enough; the picture switches to Fit and shrinks to make room, and dragging it away floats it again — double-click the grip to put it back); **More — int/ext & time of day** holds those two fields. On a phone the panel is a sheet that slides up from the bottom — tap the sliders icon in the top bar to open it
* **Takes strip** — every generation runs along the bottom as a strip of thumbnails. Click one, or use the **←** and **→** keys, to put it on screen; the take with the green tick is the active image
* **Action bar** — thumbs up/down, favourite, info, hide, delete and **Set Active** sit in a bar just above the strip; the bookmark button there saves the take as a view (see Saving Views below). The heart and eye buttons at the left of the strip filter to favourites or reveal hidden takes
* **Fill or Fit** — the top-right switch crops the image to fill the screen or fits the whole image inside it. Click the image itself to open it full size
* **Closing** — the **×** in the top right (or **Esc**) saves anything unsaved and returns you to your locations. **Switch to classic view** at the bottom of the panel goes back to the dialog

<Note>
  **Tip:** The AI model in use is shown just below the generate button — click **Click here to change** to experiment with different models for varied visual styles and quality levels.
</Note>

### Saving Views

A single location can hold several **views** — for example a wide establishing angle, a tight corner, or the same room at a different time of day. Saving a generation as a view lets you reuse that exact image as the reference when generating a shot's first frame, without changing the location's main active image.

* **Save a view** — Navigate to the generation you want to keep and click the bookmark (**+**) button beneath the image. The view is saved with an automatic name like "View 2"
* **Rename a view** — Click the view's name badge and enter a new name (e.g., "Wide establishing", "Behind the bar")
* **Remove a view** — Click the bookmark-off button to unsave it; the underlying generation is kept, it just stops appearing in the shot view picker

<Note>
  Views are per-location. The location's **active image** acts as the default view for any shot that doesn't pick a specific one. See [Generate First Frame](/tutorial/shot-studio/generate-first-frame) for how to select a view when creating a shot.
</Note>

> ⚠️ **Important:** You must create at least one location before you can create a scene.

***

<CardGroup>
  <Card title="Character Studio" icon="arrow-left" href="/tutorial/character-studio">
    Previous: create characters
  </Card>

  <Card title="Prop Studio" icon="cube" href="/tutorial/prop-studio">
    Next: create props for your scenes
  </Card>
</CardGroup>


# Prop Studio
Source: https://docs.automatstudio.com/tutorial/prop-studio

Create and manage props for your film

## Create a Prop

Props are objects that appear in your film — anything from a coffee mug to a vintage car. Each prop can be reused across multiple scenes and shots, ensuring visual consistency throughout your production.

Prefer the picture big? **Switch to fullscreen view** (bottom left of the dialog, next to Watch Tutorial) opens the same prop with the image filling the screen and the controls floating on top — see [Fullscreen view](#fullscreen-view) below. Whatever you have typed comes with you.

### Step 1: Enter Prop Details

* On the Prop Studio page, click **Create New Prop**
* Enter the prop's **Name** (e.g., "Antique Pocket Watch" or "Leather Briefcase")
* Write a detailed **Description** — include material, color, size, condition, distinctive features, and any other visual characteristics that will help the AI generate an accurate representation

### Step 2: Choose Draft or Final quality

Under the generate button you'll find a **Draft / Final** quality switch:

* **Draft** — fast and inexpensive, ideal while you're iterating on the description
* **Final** — the best image quality on offer, for when you're happy with the direction

The generate button always shows exactly what a generation will cost with the current choice. The model line below the switch shows which AI model is in use — click **Click here to change** to pick any supported model instead; the switch clears until you tap a quality again, and your quality choice carries over when you swap between generation methods.

### Step 3: Generate Prop Image

* Click **Generate Prop Image** — the details you entered will guide the AI image generation
* Review the generated image and refine your description if needed
* Generate again until you're satisfied with the result
* Click **Save and Close** to add the prop to your project

### Managing Generations

Once you've generated prop images, you can manage them with the following features:

* **Navigate through generations** — Use the back and next buttons under the image to browse through all generated versions
* **Provide feedback** — Click the thumbs up and thumbs down buttons to give feedback on generated images
* **Active image** — The active image will be used when creating shots and scenes. New generations are automatically set as active
* **Set active image** — To set a different image as active, navigate to it using the back/next buttons and click **Set Active** underneath the image
* **Hide a generation** — Click the eye button to tuck away a generation you no longer want to see, without losing it. To bring it back, click the show-hidden (eye) button next to the navigation counter, navigate to the hidden generation, and click **Unhide**
* **Delete a generation** — Click the trash button to remove a generation permanently. If you might want it back later, hide it instead

### Fullscreen view

Click **Switch to fullscreen view** at the bottom of the dialog to work with the image filling the whole screen. Everything above works the same way; the layout is different:

* **Controls panel** — Name, the New / Edit / Reference control, the description, Draft / Final and the generate button float in a panel at the left — drag it by the grip along its top edge to uncover the picture, or drag it to the left or right side of the window to dock it there as a rail (a guide appears when it is close enough; the picture switches to Fit and shrinks to make room, and dragging it away floats it again — double-click the grip to put it back). On a phone the panel is a sheet that slides up from the bottom — tap the sliders icon in the top bar to open it
* **Takes strip** — every generation runs along the bottom as a strip of thumbnails. Click one, or use the **←** and **→** keys, to put it on screen; the take with the green tick is the active image
* **Action bar** — thumbs up/down, favourite, info, hide, delete and **Set Active** sit in a bar just above the strip. The heart and eye buttons at the left of the strip filter to favourites or reveal hidden takes
* **Fill or Fit** — the top-right switch crops the image to fill the screen or fits the whole image inside it. Click the image itself to open it full size
* **Closing** — the **×** in the top right (or **Esc**) saves anything unsaved and returns you to your props. **Switch to classic view** at the bottom of the panel goes back to the dialog

<Note>
  **Tip:** The AI model in use is shown just below the generate button — click **Click here to change** to experiment with different models for varied visual styles and quality levels.
</Note>

***

<CardGroup>
  <Card title="Location Studio" icon="arrow-left" href="/tutorial/location-studio">
    Previous: create locations
  </Card>

  <Card title="Wardrobe Studio" icon="shirt" href="/tutorial/wardrobe-studio">
    Next: create costumes and outfits
  </Card>
</CardGroup>


# Share and Export Projects
Source: https://docs.automatstudio.com/tutorial/share-and-export

Share your projects, export media assets, or export an editable timeline (XML) for Final Cut Pro, DaVinci Resolve, and Premiere Pro

Once you've completed your project, you can share it with others or export your media assets for use outside of Automat Studio.

## Share Project

You can create a shareable link to share your project with collaborators, clients, or viewers.

### How to Share

1. Navigate to the **View Project** page
2. Click **Share**
3. If you're on the **Hollywood plan**, you can choose to upscale the project to 4K resolution
4. Click **Share Project** to generate the share link
5. Previous share links will appear in the list below

<Note>
  **Free tier projects** will include watermarks on shared content. Upgrade to a paid plan to share without watermarks.
</Note>

<Note>
  Shots with [in and out points](/tutorial/shot-studio/trim-in-out-points) are cut to their trimmed range before the film is stitched.
</Note>

***

## Export Assets

You can export your project's media assets if you're on the **Independent plan** or higher. This allows you to download images, videos, and audio files for use in external editing software or other platforms.

### How to Export

1. Navigate to the **View Project** page
2. Click **Export Media**
3. Select the type of media you want to export:
   * **Images** — Character and location images
   * **Videos** — Generated shot videos
   * **Audio** — Dialogue and sound effects
4. Choose your export scope:
   * **All media** — Export all generated assets
   * **Active generations only** — Export only the currently active versions
5. Click **Export Media** to start the download

Your download will begin automatically once the export is processed.

<Note>
  Export functionality is available on **Independent plan** and higher. Free tier users can upgrade to access this feature.
</Note>

<Note>
  Exported videos are the original, untrimmed files. A shot's [in and out points](/tutorial/shot-studio/trim-in-out-points) are carried by the XML timeline export instead.
</Note>

***

## Export XML (Timeline Export)

You can export your whole film as an editable timeline for professional editing software — Final Cut Pro, DaVinci Resolve, or Adobe Premiere Pro — to continue editing, sound mixing, and color grading outside Automat Studio.

The export is a single ZIP package containing:

* **project.fcpxml** — import into **Final Cut Pro** or **DaVinci Resolve**
* **project.xml** — import into **Adobe Premiere Pro** (DaVinci Resolve reads this too)
* **media/** — every shot's video (and dialogue/SFX audio) in timeline order
* **README.txt** — import steps for each editor

Your shots are assembled into a sequence in scene and shot order, with dialogue/SFX audio on a separate track and a timeline marker on every clip carrying its scene, shot number, and description. A shot's [in and out points](/tutorial/shot-studio/trim-in-out-points) become the clip's source in/out in the timeline — the media file is still the whole clip, so you can extend the cut in your editor.

### How to Export XML

1. Navigate to the **View Project** page
2. Open the **⋯** menu and click **Export XML**
3. Click **Export XML** to download the package
4. Unzip the whole package, keeping the XML files next to the `media` folder
5. Import the matching XML file into your editor (see the included README)

<Note>
  Shots without a generated video are not included in the timeline. If your editor reports offline media after import, use its relink command and point it at the `media` folder.
</Note>

<Note>
  XML Project Export is available on the **Hollywood plan**.
</Note>

***

<CardGroup>
  <Card title="Shot Studio" icon="arrow-left" href="/tutorial/shot-studio/index">
    Previous: create shots
  </Card>
</CardGroup>


# Dialogue and Lipsync
Source: https://docs.automatstudio.com/tutorial/shot-studio/dialogue-lipsync

Add voice performances that stay in sync with your characters

The Dialogue and Lipsync tab allows you to generate dialogue audio from text and automatically sync it to your character's lip movements. This two-step workflow lets you create natural-sounding voice performances and apply them to your rendered shots.

## Workflow

### Step 1: Generate Dialogue Audio

1. Enter the dialogue text you want the character to speak
2. Select from the list which character is speaking
3. Click **Generate Dialogue** — Automat will create audio using text-to-speech models
4. Listen to the generated audio clip
5. Generate multiple times — Each generation will have a different performance and delivery
6. Cycle through generations and click **Play** to listen and compare takes
7. Once you're satisfied with a take, select it to use for lipsync

<Note>
  You can make as many dialogue generations as you like. Each generation will have a slightly different performance, allowing you to find the perfect delivery for your scene. See the [Text-to-Audio models guide](/models/text-to-audio) for available voice models.
</Note>

### Step 2: Apply Lipsync to Video

1. With your preferred dialogue audio selected, click **Generate Lipsync**
2. This will apply the selected dialogue to the active video generation
3. Wait for the lipsync generation to complete (typically 30-90 seconds)
4. Watch the new video in the left-hand side viewer
5. Review the synced result — If the sync isn't perfect, adjust settings and regenerate

<Note>
  Lipsync generation applies the dialogue audio to your rendered video, automatically matching lip movements, facial expressions, and subtle head motions to the voice performance. See the [Audio-to-Video models guide](/models/audio-to-video) for details on available lipsync models.
</Note>

## Model Selection

The model in use is shown just below each generate button — "Model: *model name*". Click **Click here to change** to open the model picker. This allows you to:

* Select dialogue and lipsync models based on quality vs. speed preferences
* Experiment with different model combinations for best results

<Note>
  **Recommended models** provide the best balance of quality and speed. **Supported models** may have different characteristics—experiment to find what works best for your specific shot.
</Note>

## Tips for Best Results

1. **Generate Multiple Takes** — Create several dialogue generations to compare different performances
2. **Match Character Voices** — Use consistent voice profiles for each character across your project
3. **Check Sync Accuracy** — Review the final video to ensure lip movements match the dialogue timing
4. **Iterate if Needed** — Regenerate lipsync with different settings or models if the sync isn't perfect

***

<CardGroup>
  <Card title="Performance" icon="arrow-left" href="/tutorial/shot-studio/performance">
    Previous: performance mode
  </Card>

  <Card title="VFX (Visual Effects)" icon="sparkles" href="/tutorial/shot-studio/vfx-visual-effects">
    Next: add visual effects
  </Card>
</CardGroup>

***

<CardGroup>
  <Card title="Text-to-Audio Models" icon="microphone" href="/models/text-to-audio">
    Learn about available dialogue generation models
  </Card>

  <Card title="Audio-to-Video Models" icon="lips" href="/models/audio-to-video">
    Explore lipsync model options and capabilities
  </Card>
</CardGroup>


# Generate First Frame
Source: https://docs.automatstudio.com/tutorial/shot-studio/generate-first-frame

Create the hero still that anchors your shot

Use the Generate First Frame tab to iterate on the look and feel of the shot before moving into motion.

## Workflow

1. Update the shot prompt, lens, and description to match the story beat.
2. Click **Generate First Frame** and review the rendered still in the gallery.
3. Try multiple variations, refining prompts or camera settings between runs.
4. Save the take that best represents your shot before proceeding to video generation.

<Note>
  Each render is stored in the shot history so you can compare versions or revert at any time.
</Note>

\--

### Step 1: Enter Shot Details

* Click **Create New Shot**
* Click **Open Shot**
* Select:
  * **Lens**
  * **Shot Type**
* Add a description of the first frame in ***first frame description***
* Click **Save** to save your progress at any time

### Step 2: Choose Draft or Final quality

Under the generate button you'll find the same **Draft / Final** quality switch as the asset studios:

* **Draft** — fast and inexpensive, ideal while you're iterating on the framing
* **Final** — the best image quality on offer, for when the shot is locked

The generate button always shows exactly what a generation will cost with the current choice, and the model line below the switch shows which AI model is in use — click **Click here to change** to pick any supported model instead.

### Step 3: Generate First Frame

* Click **Generate First Frame** to create the first image for your shot
* Update the description, lens, or shot type and generate again until you’re happy

> ℹ️ Note: Shot attributes (like lens and shot type) may not be precisely reflected in the image due to current AI limitations — but this will improve over time!

### Choosing a Location View

The shot's location is set by its scene, and by default the location's active image is used as the reference for the first frame. If you've saved multiple [views](../location-studio#saving-views) for that location, you can choose which one to use:

* Enable **Show Advanced Options** to reveal the **Location view** selector
* Pick a saved view, or leave it on **Default (location active image)**

This only changes the reference for this shot, so different shots in the same scene can draw from different views of the location.

***

<CardGroup>
  <Card title="Shot Studio Overview" icon="arrow-left" href="/tutorial/shot-studio/index">
    Previous: shot studio overview
  </Card>

  <Card title="Generate Video" icon="video" href="/tutorial/shot-studio/generate-video">
    Next: animate your frame
  </Card>
</CardGroup>


# Generate Video
Source: https://docs.automatstudio.com/tutorial/shot-studio/generate-video

Turn your first frame into a fully animated shot

Once your still looks right, the Generate Video tab brings it to life with motion and timing controls.

<Note>
  Video generation is available on every plan (including free accounts) and is paid for with [credits](/support/usage), like image generation.
</Note>

## Workflow

1. Confirm the correct first frame is selected in the shot gallery
2. Enter a **video description** describing the motion and action you want to see
3. Select the **shot duration** to match your storyboard timing
4. Select the **camera movement** (static, pan, tilt, dolly, etc.) that fits your shot
5. Pick **Draft** or **Final** quality on the switch under the generate button — Draft for fast, inexpensive iteration; Final for the best motion quality when the shot is locked. The button always shows the exact credit cost for your current duration and quality, and **Click here to change** on the model line still opens the full model picker
6. Click **Generate Video** and let Automat render the clip (usually 2–3 minutes)
7. Review the playback, tweak settings if needed, and regenerate until satisfied

<Note>
  You can navigate away from the page while rendering—Automat will keep processing in the background. Changed your mind? Click **Cancel** under the progress indicator to stop the generation and get your credits refunded.
</Note>

<Note>
  **Camera motion precision:** Precise camera motion is not guaranteed, but usually pretty accurate. Try different models to get different results—each model handles camera movement and motion differently. See the [Image-to-Video models guide](/models/image-to-video) for details on each model's capabilities.
</Note>

## Tips for Better Results

* **Be specific in your video description** — Describe the type of motion, character actions, and environmental dynamics you want
* **Experiment with models** — Different models excel at different types of motion; try multiple models to find the best fit
* **Adjust duration carefully** — Shorter clips often render faster and with better quality
* **Iterate on settings** — If motion isn't right, adjust the description, duration, or camera movement and regenerate

***

<CardGroup>
  <Card title="Generate First Frame" icon="arrow-left" href="/tutorial/shot-studio/generate-first-frame">
    Previous: create first frame
  </Card>

  <Card title="Dialogue and Lipsync" icon="microphone" href="/tutorial/shot-studio/dialogue-lipsync">
    Next: add dialogue
  </Card>
</CardGroup>


# Generate Video (Direct)
Source: https://docs.automatstudio.com/tutorial/shot-studio/generate-video-direct

Skip the first frame step and generate video directly from text descriptions

The Generate Video Direct tab allows you to bypass the first frame generation step and create motion clips directly from text descriptions. This workflow is ideal when you want to quickly generate video without iterating on a first frame first.

<Note>
  Video generation is available on every plan (including free accounts) and is paid for with [credits](/support/usage), like image generation.
</Note>

<Note>
  **How to access:** To use Generate Video Direct, select the **"Direct Method"** switch in the bottom left corner of the Shot Studio interface.
</Note>

## When to Use

* You want to skip the first frame iteration step
* You need to quickly test motion concepts
* You prefer to work directly from text descriptions rather than iterating on still images

## Workflow

1. Enter a **video description** describing the motion and action you want to see
2. Select the **lens** that matches your desired shot composition
3. Select the **shot type** (close-up, medium, wide, etc.)
4. Select the **shot duration** to match your storyboard timing
5. Select the **camera movement** (static, pan, tilt, dolly, etc.) that fits your shot
6. Click **Generate Video Direct** and let Automat render the clip (typically 1-3 minutes)
7. Review the playback, adjust settings if needed, and regenerate until satisfied

<Note>
  Generate Video Direct uses image-to-video models. See the [Image-to-Video models guide](/models/image-to-video) for details on available models and their capabilities.
</Note>

## Tips for Best Results

1. **Be Specific** — Describe the type of motion, character actions, and camera movement you want
2. **Experiment with Models** — Different models excel at different types of motion; try multiple models to find the best fit
3. **Adjust Duration Carefully** — Shorter clips often render faster and with better quality
4. **Iterate on Settings** — If motion isn't right, adjust the description, duration, or camera movement and regenerate
5. **Compare with Standard Method** — Use Generate Video (standard) when you need precise first frame control

> 💡 **Tip:** This method is faster than generating a first frame separately, but you have less control over the exact starting composition. Use Generate Video (standard) when you need to fine-tune the first frame before animating.

***

<CardGroup>
  <Card title="Shot Studio Overview" icon="arrow-left" href="/tutorial/shot-studio/index">
    Previous: shot studio overview
  </Card>

  <Card title="Performance" icon="masks-theater" href="/tutorial/shot-studio/performance">
    Next: film it yourself with Performance mode
  </Card>
</CardGroup>


# Introduction
Source: https://docs.automatstudio.com/tutorial/shot-studio/index

Generate first frames, video, dialogue, SFX, and VFX for every shot

<img alt="Shot Studio overview" />

<Note>
  The [Assistant](/tutorial/assistant) can write a shot for you — description, @mentions, dialogue and speaker — and offer its first frame and video at the quoted price.
</Note>

<CardGroup>
  <Card title="Mention Characters & Props" icon="at" href="/tutorial/shot-studio/mention-assets">
    Type @ to pull your characters and props into a shot
  </Card>

  <Card title="Generate First Frame" icon="image" href="/tutorial/shot-studio/generate-first-frame">
    Create the hero still that anchors your shot
  </Card>

  <Card title="Generate Video" icon="video" href="/tutorial/shot-studio/generate-video">
    Turn your first frame into a fully animated shot
  </Card>

  <Card title="Generate Video (Direct)" icon="bolt" href="/tutorial/shot-studio/generate-video-direct">
    Skip to video generation without a first frame
  </Card>

  <Card title="Trim a Shot" icon="scissors" href="/tutorial/shot-studio/trim-in-out-points">
    Cut a generated clip down with in and out points
  </Card>

  <Card title="Performance" icon="masks-theater" href="/tutorial/shot-studio/performance">
    Film real actors and re-render the shot in any world
  </Card>

  <Card title="Dialogue and Lipsync" icon="microphone" href="/tutorial/shot-studio/dialogue-lipsync">
    Add voice performances that stay in sync with your characters
  </Card>

  <Card title="VFX (Visual Effects)" icon="sparkles" href="/tutorial/shot-studio/vfx-visual-effects">
    Layer cinematic effects onto your rendered shot
  </Card>

  <Card title="SFX (Sound Effects)" icon="volume-high" href="/tutorial/shot-studio/sfx-sound-effects">
    Design soundscapes that match your shot
  </Card>
</CardGroup>


# Mention Characters, Props & Wardrobe
Source: https://docs.automatstudio.com/tutorial/shot-studio/mention-assets

Type @ to pull your characters, props, and wardrobe into a shot for consistent results

When you describe a shot, you can reference the characters, props, and wardrobe you have already built by **mentioning** them with `@`. Mentioned assets are pulled into generation as reference images, so they appear consistently across every shot instead of being reinvented each time.

Mentions work in every shot description field — **First Frame Description**, **Edit Description**, **Video Description**, and the **VFX** and **SFX** tabs.

## Add a mention

1. Start typing in any shot description field.
2. Type `@` to open the mention picker. It lists the project's **characters** first, then **props**, then **wardrobe**, each with a thumbnail and name.
3. Keep typing to filter the list by name.
4. Choose an asset:
   * **Click** it, or
   * Use the **arrow keys** to highlight one and press **Enter** or **Tab**.
   * Press **Escape** to dismiss the picker without inserting.

The asset's name is inserted at your cursor as `@Name`, and you can keep writing your description around it.

<Note>
  The mention picker only shows assets that belong to the current project. Build a character in **Character Studio**, a prop in **Prop Studio**, or an outfit in **Wardrobe Studio** first, and it becomes available to mention.
</Note>

## How mentions are matched

* Matching is **case-insensitive** — `@alex`, `@Alex`, and `@ALEX` all resolve to the same character.
* Names that **contain spaces** are supported (for example `@Red Sports Car`).
* When two assets share a prefix, the **longest matching name wins**, so `@Red Sports Car` is preferred over `@Red`.
* A stray `@` that doesn't match an asset — such as an email address or "@ 3pm" — is left untouched and treated as ordinary text.

## What you'll see

Below the description, every resolved mention appears as a badge so you can confirm what the shot will include at a glance:

* **Characters** are shown as blue badges.
* **Props** are shown as violet badges.
* **Wardrobe** items are shown as pink badges.

If a badge you expect is missing, check the spelling against the asset's name in Character, Prop, or Wardrobe Studio.

## What mentions do at generation time

When you generate, Automat Studio gathers the **reference images** for each mentioned character, prop, and wardrobe item and supplies them to the model alongside your prompt. This is what keeps a character's face, a prop's design, or an outfit's look consistent from shot to shot.

The `@` symbols themselves are removed from the text that's sent to the model, so your prompt reads naturally — the mentions act as instructions to the studio, not literal text in the prompt.

<Note>
  Mention the characters, props, and wardrobe you actually want in the frame. Adding a mention pulls that asset's look into the result, while leaving one out keeps it out of the shot.
</Note>

***

<CardGroup>
  <Card title="Shot Studio Overview" icon="arrow-left" href="/tutorial/shot-studio/index">
    Previous: shot studio overview
  </Card>

  <Card title="Generate First Frame" icon="image" href="/tutorial/shot-studio/generate-first-frame">
    Next: create the hero still for your shot
  </Card>
</CardGroup>


# Performance
Source: https://docs.automatstudio.com/tutorial/shot-studio/performance

Film real actors, restyle the scene with AI, and keep the human performance in your shot

Performance mode lets you film real actors performing a shot, then re-render it in any world you can imagine — the actors' movement, timing, blocking and expressions are preserved, while the set, wardrobe and grade come from your project's assets. It's AI filmmaking with human acting kept fully in the loop.

<Note>
  Performance generation is available on every plan (including free accounts) and is paid for with [credits](/support/usage), like all other generation.
</Note>

## How It Works

1. **Film a take** — Record your actors performing the shot on any camera (a phone is fine). Keep it between **2 and 30 seconds**, MP4 or MOV.
2. **Upload it** — In Shot Studio's **First Frame** tab, switch to **Performance** mode and upload the take. Automat extracts the first frame automatically.
3. **Restyle the first frame** — Describe the world you want in the Restyle Description, using `@` mentions to bring in your [characters, props and wardrobe](/tutorial/shot-studio/mention-assets). The actors' poses and framing are preserved; the environment, wardrobe and lighting are repainted.
4. **Generate the video** — In the **Video** tab, switch to **Performance**. Your take drives the motion and your styled first frame drives the look. Duration is matched to your take automatically, and the take's original audio — your actors' real voices — can be kept with the **Keep my take's audio** toggle. The video description is optional here: use it for extra scene context, not choreography.

<Note>
  **How to access:** Performance mode lives in two places — the **Performance** switch in the First Frame tab (upload + restyle) and the **Performance** switch in the Video tab (final video generation).
</Note>

## When to Use

* You want real, directed human acting in an AI-generated world
* Your shot needs precise timing, blocking or emotional nuance that text prompts can't reliably produce
* You want to preview a scene in costume and location you don't have access to

## Tips for Best Results

1. **Frame it like the final shot** — The extracted first frame sets the composition, so film with the framing you want in the finished clip.
2. **Keep takes tight** — Credits scale with duration; a focused 5–8 second take usually beats a long one.
3. **Light evenly** — Clean, even lighting on the actors gives the restyle step the best chance of preserving likeness.
4. **Restyle in steps** — If a big transformation drifts, generate the restyled frame a few times or simplify the changes.
5. **Skip the restyle when you don't need it** — With no restyled frame set as active, the video is generated from the take's own first frame and your description.
6. **Keep the real voices** — With "Keep my take's audio" on, your actors' actual dialogue delivery carries into the finished shot — often the most powerful part of the workflow.
7. **Use footage you have rights to** — Upload only takes of yourself or actors who've consented.

<Note>
  Performance video uses Kling 3.0 Motion — a motion-control model built for transferring real performances onto styled frames. See the [Performance-to-Video models guide](/models/performance-to-video) for pricing and options.
</Note>

***

<CardGroup>
  <Card title="Generate Video (Direct)" icon="arrow-left" href="/tutorial/shot-studio/generate-video-direct">
    Previous: direct video generation
  </Card>

  <Card title="Dialogue and Lipsync" icon="microphone" href="/tutorial/shot-studio/dialogue-lipsync">
    Next: add dialogue and lipsync
  </Card>
</CardGroup>


# SFX (Sound Effects)
Source: https://docs.automatstudio.com/tutorial/shot-studio/sfx-sound-effects

Design soundscapes that match your shot

The SFX tab automatically generates sound effects and ambient audio that sync to the visual action in your shots. AI analyzes your video footage and creates matching sound effects, environmental audio, and foley sounds without requiring manual sound design.

<Note>
  SFX generation is available on every plan (including free accounts) and is paid for with [credits](/support/usage), like image and video generation.
</Note>

## Workflow

1. Ensure you have an active video generation selected (the SFX tab will use the current active video)
2. Optionally enter a **description** to guide the SFX generation (e.g., "subtle ambient sounds," "dramatic impact sounds," "urban street noise")
3. Select different models if available — Different models may produce different styles of sound effects
4. Click **Generate SFX** and wait for processing to complete (typically 15-30 seconds)
5. Review the generated audio — Listen to ensure it matches the visual action
6. Adjust and regenerate if needed — Try different descriptions or models to refine the sound design

<Note>
  SFX generation uses video-to-audio models that analyze your video and automatically generate matching sound effects. See the [Video-to-Audio models guide](/models/video-to-audio) for details on available models and their capabilities.
</Note>

<Note>
  You can always replace or layer additional audio tracks later if you need to make adjustments or add director-requested changes.
</Note>

## Example SFX Descriptions

Here are examples of effective SFX descriptions to guide generation:

* **Ambient sounds:** "subtle ambient room tone," "forest atmosphere," "urban street ambience"
* **Action-focused:** "dramatic impact sounds," "quick footsteps on concrete," "door slam and echo"
* **Environmental:** "wind rustling leaves," "distant traffic," "ocean waves on shore"
* **Foley effects:** "footsteps on gravel," "paper rustling," "metal clinking"
* **Atmospheric:** "tension-building ambient," "mysterious background sounds," "energetic soundscape"

## Tips for Best Results

1. **Use Active Video** — Make sure the video generation you want SFX for is currently active/selected
2. **Be Descriptive** — Include specific details about the type of sounds you want (ambient, action, environmental)
3. **Try Different Models** — Experiment with different models to find the best sound style for your shot
4. **Review Carefully** — Listen to generated SFX to ensure it matches the visual timing and mood

> 💡 **Tip:** Start with a general description, then refine with more specific details if the initial generation isn't quite right. You can always regenerate with different prompts.

***

<CardGroup>
  <Card title="VFX (Visual Effects)" icon="arrow-left" href="/tutorial/shot-studio/vfx-visual-effects">
    Previous: add visual effects
  </Card>

  <Card title="Share and Export" icon="share" href="../share-and-export">
    Next: share and export your project
  </Card>
</CardGroup>

***

<CardGroup>
  <Card title="Video-to-Audio Models" icon="waveform" href="/models/video-to-audio">
    Learn about available SFX generation models
  </Card>
</CardGroup>


# Trim a Shot (In and Out Points)
Source: https://docs.automatstudio.com/tutorial/shot-studio/trim-in-out-points

Cut a generated clip down to the part you want with in and out points

A generated clip is a fixed length, but the take you keep is often shorter — a slow start, a wobble at the end, or a beat that should land tighter. In and out points let you cut the clip down without regenerating it.

The trim controls sit under the video preview in the Shot Studio, and appear once a take is the shot's **active video**.

## Three ways to set the points

* **Drag the handles.** The bar under the player shows the whole clip. Drag the left handle to set the in point and the right handle to set the out point. The preview scrubs to the frame under the handle as you drag, and the trim saves when you let go.
* **Mark at the playhead.** Scrub the player to the exact frame, then click **Set In** or **Set Out** — or press <kbd>I</kbd> or <kbd>O</kbd>, as in Premiere and Final Cut.
* **Type a timecode.** Click either timecode to edit it. Times are `SS:FF` (seconds and frames at the clip's frame rate), so `01:12` is one second and twelve frames. Press <kbd>Enter</kbd> to apply or <kbd>Esc</kbd> to cancel.

With a handle focused, the arrow keys move it one frame and <kbd>Shift</kbd> + arrow moves it ten. **Clear** removes the trim.

<Note>
  The preview plays only the trimmed range: pressing play starts at the in point and playback stops at the out point. Scrub outside the range to review what you cut.
</Note>

## What the trim applies to

The trim belongs to the shot and describes its active video. Where it takes effect:

* **Shot Studio preview** and the **scene and project players**
* **Runtimes** on shot, scene and project cards
* **Share Project** — the shared film is stitched from the trimmed clips
* **Export XML** — the Final Cut Pro, DaVinci Resolve and Premiere Pro timelines carry the in and out points as a source trim, so the whole clip is still available in your editor

**Export Media** downloads the original, untrimmed files.

<Note>
  Choosing a different take as the active video, or generating a new one, clears the trim — a new take always starts untrimmed. Duplicating a shot copies its trim.
</Note>

***

<CardGroup>
  <Card title="Generate Video" icon="video" href="/tutorial/shot-studio/generate-video">
    Turn your first frame into a fully animated shot
  </Card>

  <Card title="Share and Export" icon="share-nodes" href="/tutorial/share-and-export">
    Share the film or export a timeline for your editor
  </Card>
</CardGroup>


# VFX (Visual Effects)
Source: https://docs.automatstudio.com/tutorial/shot-studio/vfx-visual-effects

Layer cinematic effects onto your rendered shot

The VFX tab allows you to enhance your rendered shots by adding visual effects, making style changes, or modifying elements within the frame. Apply everything from subtle lighting adjustments to dramatic visual transformations without regenerating the entire shot.

<Note>
  VFX generation is available on every plan (including free accounts) and is paid for with [credits](/support/usage), like image and video generation.
</Note>

## What You Can Do

The VFX tab supports three levels of visual modifications:

### Subtle Changes

* **Lighting adjustments** — Change time of day, add golden hour lighting, adjust shadows
* **Color grading** — Shift color palettes, increase saturation, apply color filters
* **Atmospheric effects** — Add fog, rain, snow, or dust particles
* **Depth and focus** — Adjust depth of field, add lens blur effects

### Adding Elements

* **Environmental elements** — Add clouds, stars, rain, snow, or other weather
* **Objects and props** — Insert objects, characters, or background elements
* **Particle effects** — Add sparks, smoke, magical effects, or debris
* **Lighting elements** — Add light sources, lens flares, or glowing effects

### Complete Transformations

* **Style changes** — Transform entire scenes (e.g., "make it nighttime," "add film grain")
* **Costume/prop changes** — Modify character appearance (e.g., "change her suit to yellow")
* **Environmental overhauls** — Completely change settings or add dramatic elements (e.g., "add a tentacled alien floating in the sky")

## Workflow

1. Select the video generation you want to enhance from your shot gallery
2. Enter a **VFX description** describing the visual effect or change you want to apply
   re
3. Click **Generate VFX** and wait for processing to complete (typically 2-3 minutes)
4. Review the enhanced video in the viewer
5. If satisfied, save the result; if not, adjust your description and regenerate

<Note>
  VFX passes are non-destructive—you can revert to the original video at any time. Each VFX generation creates a new version without overwriting your original shot.
</Note>

<Note>
  VFX generation uses video-to-video models. See the [Video-to-Video models guide](/models/video-to-video) for details on available models and their capabilities.
</Note>

## Example VFX Prompts

Here are examples of effective VFX descriptions:

* **Time of day:** "make it nighttime," "add golden hour lighting," "change to sunset"
* **Weather:** "add heavy rain," "make it snowing," "add fog rolling in"
* **Style changes:** "add film grain," "make it look like a vintage photograph," "apply a cyberpunk color palette"
* **Object modifications:** "change her suit to yellow," "add sunglasses to the character," "replace the car with a motorcycle"
* **Environmental additions:** "add a tentacled alien floating in the sky," "add birds flying in the background," "insert a spaceship in the distance"
* **Lighting effects:** "add dramatic shadows," "increase ambient lighting," "add a spotlight from above"
* **Atmospheric effects:** "add dust particles in the air," "create a dreamy haze," "add lens flares"

## Tips for Best Results

1. **Be Specific** — Clearly describe the visual change you want (e.g., "change suit to bright yellow" vs. "change colors")
2. **Start Subtle** — Test subtle effects first before attempting major transformations
3. **Iterate Carefully** — VFX passes build on each other; review results before applying multiple effects
4. **Match Project Style** — Ensure VFX effects align with your project's overall aesthetic
5. **Use Precise Language** — Specify exact colors, locations, and intensity levels when possible

> 💡 **Tip:** VFX passes are non-destructive—you can always revert to the original render if the effect doesn't work as intended. Experiment freely!

***

<CardGroup>
  <Card title="Dialogue and Lipsync" icon="arrow-left" href="/tutorial/shot-studio/dialogue-lipsync">
    Previous: add dialogue
  </Card>

  <Card title="SFX (Sound Effects)" icon="volume-high" href="/tutorial/shot-studio/sfx-sound-effects">
    Next: add sound effects
  </Card>
</CardGroup>

***

<CardGroup>
  <Card title="Video-to-Video Models" icon="wand-magic-sparkles" href="/models/video-to-video">
    Learn about available VFX models and capabilities
  </Card>
</CardGroup>


# Wardrobe Studio
Source: https://docs.automatstudio.com/tutorial/wardrobe-studio

Create and manage wardrobe — costumes and outfits — for your film

## Create a Wardrobe Item

Wardrobe items are the costumes and outfits in your film — anything from a tailored tuxedo to a weathered leather jacket. Like props, each wardrobe item can be reused across multiple scenes and shots, keeping a look consistent throughout your production.

Prefer the picture big? **Switch to fullscreen view** (bottom left of the dialog, next to Watch Tutorial) opens the same wardrobe item with the image filling the screen and the controls floating on top — see [Fullscreen view](#fullscreen-view) below. Whatever you have typed comes with you.

### Step 1: Enter Wardrobe Details

* On the Wardrobe Studio page, click **Create New Wardrobe**
* Enter the wardrobe item's **Name** (e.g., "Detective's Trench Coat" or "Red Gala Dress")
* Write a detailed **Description** — include fabric, color, cut, fit, era, condition, and any distinctive details (buttons, embroidery, wear) that will help the AI generate an accurate costume

### Step 2: Link a Character (optional)

* Use the optional **Character** dropdown to associate the outfit with one of your project's characters
* This is purely for **organization** — grouping outfits under the character who wears them. It does **not** change how the image is generated, and it doesn't tie the wardrobe to that character in shots
* Leave it as **None** if the outfit isn't tied to a specific character

### Step 3: Choose Draft or Final quality

Under the generate button you'll find a **Draft / Final** quality switch:

* **Draft** — fast and inexpensive, ideal while you're iterating on the description
* **Final** — the best image quality on offer, for when you're happy with the direction

The generate button always shows exactly what a generation will cost with the current choice. The model line below the switch shows which AI model is in use — click **Click here to change** to pick any supported model instead; the switch clears until you tap a quality again, and your quality choice carries over when you swap between generation methods.

### Step 4: Generate Wardrobe Image

* Click **Generate Wardrobe Image** — the details you entered will guide the AI image generation
* The outfit is rendered on its own (on an invisible mannequin) so the garment's shape and details are clear, rather than worn by a person
* Review the generated image and refine your description if needed
* Generate again until you're satisfied with the result
* Click **Save and Close** to add the wardrobe item to your project

### Managing Generations

Once you've generated wardrobe images, you can manage them with the following features:

* **Navigate through generations** — Use the back and next buttons under the image to browse through all generated versions
* **Provide feedback** — Click the thumbs up and thumbs down buttons to give feedback on generated images
* **Active image** — The active image will be used when creating shots and scenes. New generations are automatically set as active
* **Set active image** — To set a different image as active, navigate to it using the back/next buttons and click **Set Active** underneath the image
* **Hide a generation** — Click the eye button to tuck away a generation you no longer want to see, without losing it. To bring it back, click the show-hidden (eye) button next to the navigation counter, navigate to the hidden generation, and click **Unhide**
* **Delete a generation** — Click the trash button to remove a generation permanently. If you might want it back later, hide it instead

<Note>
  **Using wardrobe in a shot:** mention a wardrobe item with `@` in any shot description (see [Mention Characters, Props & Wardrobe](/tutorial/shot-studio/mention-assets)). In this version the outfit is supplied to the model as a reference image alongside your characters — describe in the shot how you want it worn.
</Note>

### Fullscreen view

Click **Switch to fullscreen view** at the bottom of the dialog to work with the image filling the whole screen. Everything above works the same way; the layout is different:

* **Controls panel** — Name, the New / Edit / Reference control, the description, Draft / Final and the generate button float in a panel at the left — drag it by the grip along its top edge to uncover the picture, or drag it to the left or right side of the window to dock it there as a rail (a guide appears when it is close enough; the picture switches to Fit and shrinks to make room, and dragging it away floats it again — double-click the grip to put it back); the optional character link sits under the name. On a phone the panel is a sheet that slides up from the bottom — tap the sliders icon in the top bar to open it
* **Takes strip** — every generation runs along the bottom as a strip of thumbnails. Click one, or use the **←** and **→** keys, to put it on screen; the take with the green tick is the active image
* **Action bar** — thumbs up/down, favourite, info, hide, delete and **Set Active** sit in a bar just above the strip. The heart and eye buttons at the left of the strip filter to favourites or reveal hidden takes
* **Fill or Fit** — the top-right switch crops the image to fill the screen or fits the whole image inside it. Click the image itself to open it full size
* **Closing** — the **×** in the top right (or **Esc**) saves anything unsaved and returns you to your wardrobe. **Switch to classic view** at the bottom of the panel goes back to the dialog

<Note>
  **Tip:** The AI model in use is shown just below the generate button — click **Click here to change** to experiment with different models for varied visual styles and quality levels.
</Note>

***

<CardGroup>
  <Card title="Prop Studio" icon="arrow-left" href="/tutorial/prop-studio">
    Previous: create props
  </Card>

  <Card title="Create a Scene" icon="film" href="/tutorial/create-a-scene">
    Next: combine everything into scenes
  </Card>
</CardGroup>