# moviefactory.ai > Join early access to MovieFactory.ai, an AI film studio for turning your script into characters, worlds, storyboards and films. Canonical website: https://moviefactory-ai.vercel.app/ MovieFactory.ai is a browser-based AI filmmaking workspace. MovieFactory.ai is in early access. Join the beta waitlist on the website to register your interest. Invited users connect their own provider accounts. ## Scope and limitations The website describes the product and its production workflow. Writing, image, video and audio generation takes place in the studio using connected providers. Generation requires eligible provider access and may consume provider credits. Camera, lens and style controls guide generation; results depend on the selected model. Projects and uploaded media stay in the current browser on this device. Export backups to keep them; there is no automatic cloud synchronization. AI results depend on the model, inputs and direction, and must be reviewed. ## Product pages - [Website](https://moviefactory-ai.vercel.app/): Product overview and production workflow. - [Production workflow](https://moviefactory-ai.vercel.app/#studio): Write, direct, generate and edit. - [Join early access](https://moviefactory-ai.vercel.app/#early-access): Register your interest in the MovieFactory.ai beta. - [Frequently asked questions](https://moviefactory-ai.vercel.app/#questions): Pricing, storage, media and export. ## What you can create ### Full films Build a narrative scene by scene. Keep recurring characters, locations and shot direction together as you work toward a complete cut. Deliverable: A finished film with picture and sound. ### Storyboards & animatics Test composition, shot order and pacing with still frames. Export an animatic before committing to video generation or a physical shoot. Deliverable: A timed visual plan for your production. ### Trailers & concept films Explore the mood of a bigger story. Combine a few directed shots with a soundtrack to show the world, tension and visual language. Deliverable: A concise proof of concept. ### Creative experiments Try another model, revise a camera move or compare alternate frames. Keep earlier takes available while you decide what works. Deliverable: A collection of ideas you can compare. ## From script to screen ### 01. Paste or upload your script. Start with the story you already have. Paste your screenplay or upload a file, then analyze it to create a proposed cast, world references and shot plan. Review the breakdown before you apply it. - Fountain or plain text - An editable scene and shot plan ### 02. Give your characters a face. Review the characters from your script and refine their look. Generate reference images, try another take or upload your own. Choose the images that feel right and assign your cast to each shot. - Generate or upload references - Reuse your cast across scenes ### 03. Build the world around them. Create the locations, environments and props your story needs. Describe a room, a landscape or an imagined world, generate its visual references and use them wherever the story returns. - Locations, environments and props - Generate missing references together ### 04. Make each shot your own. Bring your cast and locations into the shot plan. Prepare reference frames and choose the framing and movement. Go deeper with lenses, camera angles and visual styles when you want more control. - Cast and world references per shot - Advanced camera and lens controls ### 05. Bring your scenes to life. Choose a connected video model and generate a take from your shot direction. Compare versions, keep the one you love and create voice or sound to support the scene. - Your choice of video model - Saved takes to compare and select ### 06. Bring it all together. Arrange your clips and stills in the timeline, adjust the pace and add your soundtrack. Export a movie you can share and save a production backup to pick up where you left off. - Picture and sound in one timeline - A finished movie to share ## Production tools - Screenplay & breakdown: Paste a script or import Fountain or plain text. Review the proposed cast, locations and shots before applying the breakdown. - Cast & world library: Generate or upload character, location and environment references. Keep the versions you choose connected to your shots. - First & last frames: Set endpoints and direct each shot. Advanced mode adds lens type, focal length, shot size, camera angle, movement, focus and visual style to your frame and video prompts. - Take comparisons: Compare saved generations and choose which version belongs in the edit. - Timeline & soundtrack: Sequence clips and stills, adjust the pace and add a film-wide soundtrack. - Movie & project export: Export your movie, download your screenplay and back up a portable production package. ## Models and connections - [GPT Image 2](https://developers.openai.com/api/docs/models/gpt-image-2) (OpenAI; Built-in connection): Create character references, locations and opening or closing frames. Generate or edit images using your selected cast and world references. - [Veo 3.1](https://docs.dev.runwayml.com/guides/models/) (Google · via Runway or Higgsfield; Built-in connection): Turn first and last frames into a directed shot with optional audio. Standard and Fast modes. Choose landscape or portrait and review each take before adding it to the cut. - [Veo 3.1 Fast](https://docs.dev.runwayml.com/guides/models/) (Google · via Runway or Higgsfield; Built-in connection): Explore a shot with the Fast model while keeping your frame direction. Use first and last images, optional audio and clips up to eight seconds. Save and compare takes before choosing your cut. - [LTX 2.5](https://docs.comfy.org/tutorials/video/ltx/ltx-2-5) (Lightricks · via ComfyUI; ComfyUI preset): Guide a shot with opening and closing images and generate picture with sound. The distilled preset checks your installed models and nodes before setup. - [MiniMax H3](https://docs.comfy.org/tutorials/video/minimax/minimax-h3) (MiniMax · via ComfyUI; ComfyUI preset): Generate motion and scene sound in a joint video and audio workflow. The turbo preset requires compatible model files and nodes. Commercial use of local outputs requires a MiniMax commercial license. - [GPT-5 mini](https://developers.openai.com/api/docs/models/gpt-5-mini) (OpenAI; Built-in connection): Develop a Fountain screenplay and turn a script into an editable shot plan. The default text model is configurable. Review drafts and analysis before applying them. - [Eleven Multilingual v2](https://elevenlabs.io/docs/overview/models) (ElevenLabs; Built-in connection): Give dialogue and narration a voice from your connected voice library. Choose a voice, edit the spoken text and review the generated audio. - [Sound Effects v2](https://elevenlabs.io/docs/overview/capabilities/sound-effects) (ElevenLabs; Built-in connection): Build ambience, transitions and scene effects from a sound direction. Generate effects up to 30 seconds. Speech duration follows the text. - [ComfyUI](https://docs.comfy.org/) (Your server · your workflows; Built-in connection): Bring your own image, video, music or sound generation pipeline. Import an API-format workflow and map its prompt, frames, duration and output inputs. - [Runway](https://docs.dev.runwayml.com/) (Video API; Built-in connection): Send Veo 3.1 and Veo 3.1 Fast shots directly from the render studio. The current connection supports Veo models; other Runway models are not exposed in this studio. - [Higgsfield](https://docs.higgsfield.ai/) (Video API; Built-in connection): Create Veo 3.1 shots with first- and last-frame direction. Connect a key ID and secret. This integration uses standard or fast Veo endpoints. - [FLUX](https://docs.comfy.org/tutorials/flux/flux-1-text-to-image) (Black Forest Labs · via ComfyUI; Custom workflow): Explore another image workflow for concept art and visual references. Bring a compatible ComfyUI API workflow with its models and nodes installed. No built-in FLUX preset. - [Qwen-Image](https://docs.comfy.org/tutorials/image/qwen/qwen-image) (Qwen · via ComfyUI; Custom workflow): Use a custom image workflow to develop the visual world of your film. Requires a mapped ComfyUI API workflow. Compatibility depends on the workflow and installed nodes. - [Wan 2.2](https://docs.comfy.org/tutorials/video/wan/wan2_2) (Wan · via ComfyUI; Custom workflow): Bring an alternative video pipeline into your production process. Requires your own compatible workflow, models and input mappings. No automatic Wan preset is included. ## Product questions ### How do I join the early beta? Add your email to the early-access list on this page. We’ll use it to share news about MovieFactory.ai beta access. Signing up registers your interest; invitations will be shared separately. ### What is moviefactory.ai? moviefactory.ai is a browser-based AI filmmaking workspace in early access. It connects screenplay writing, storyboarding, cast and world references, shot direction, AI generation, sound and editing in one production project. ### What can I create with moviefactory.ai? Build full films, cinematic scenes, storyboards, animatics, trailers and concept films. Start from a story or screenplay, or upload your own images, clips and sound to assemble an edit. ### Can I start with my own script? Yes. Paste your screenplay into the Script room or use Import script to upload a .fountain or .txt file up to 2 MB. Analyze the script to propose characters, locations, environments, props and shots, then review the breakdown before applying it. You can generate the character and world references when you apply the plan, or create them individually later. ### How do I create characters and locations? Start with the references proposed from your script, or add your own. Refine a character’s appearance or describe a location, environment or prop, then generate an image. You can also upload PNG, JPEG or WebP references, generate missing images together and choose new takes. Assign your selected references to shots to guide the look of the film. Generation uses your connected provider account. ### Do I need to be an experienced filmmaker? You can start with your script and follow the production steps: review the breakdown, create your cast and locations, prepare shots, generate takes and assemble the edit. Refine one scene at a time, with advanced camera, lens and style controls available when you want them. ### Which AI models and tools can I connect? Built-in connections include OpenAI for writing and GPT Image 2; Runway and Higgsfield for Veo 3.1 and Veo 3.1 Fast; ElevenLabs for speech and sound effects; and ComfyUI for custom generation. ComfyUI presets support MiniMax H3 turbo and LTX 2.5 distilled when their required models and nodes are installed. Other models such as FLUX, Qwen-Image and Wan need a compatible custom ComfyUI workflow. ### Does every listed model work automatically? No. Built-in connections need an eligible provider account and credentials. ComfyUI presets require the matching models and nodes on your server. Custom-workflow entries are examples of ComfyUI pipelines you can configure, not preinstalled or individually verified integrations. Model access, cost, hardware and workflow compatibility vary. ### How does AI generation pricing work? Generation uses the connected provider account and its credits, or your own ComfyUI hardware. Costs and available models depend on the provider and settings you choose. Provider API access and credits may be separate from consumer subscriptions. ### Can I use my own clips, images and audio? Yes. Upload your own media, reuse images as shot endpoints, combine still frames with video clips, and select a soundtrack. You can build an animatic with stills before generating motion. ### How do I keep characters consistent across shots? Create shared cast and world references, assign them to shots and reuse selected images when preparing new frames. You can also reuse a shot’s closing frame as the next opening frame. These controls guide continuity; the final results depend on the model and direction, so review your takes. ### Where are my projects saved? Projects and uploaded media are saved in this browser on this device. They are not automatically synced to the cloud. Export a production package to back up your editable project and local media, or to move it to another browser. Provider keys you enter are kept for the browser session and are excluded from project exports. ### Can I export a finished movie? Yes. Assemble your clips, still frames and soundtrack, then export your finished movie. Export runs in your browser, so keep the tab active until it finishes. You can also download screenplays, an edit decision list and your editable production package. Provider and model names identify their respective owners. A listing is not a partnership or endorsement. Custom workflows are not guaranteed compatible.