Framo Documentation
Guides and reference for Framo — the AI-native design workspace for generating and editing images, video, and 3D models on an infinite canvas.
Written for people and AI assistants alike. Machine-readable index at /llms.txt.
Concepts
The big-picture mental models — surfaces, how generation works, billing.
- Plans, Credits & OrganizationsHow credits, plans, top-ups, and organization (team) billing work.
- Choosing a ModelMatch a task — image, edit, video, 3D, upscale, or training — to the right model family, grounded in the live model registry.
- Getting StartedThe big picture — surfaces (Scene / AI canvas / chat), how generation works, and where to get help.
Features
How each part of Framo works, surface by surface.
- The Agent (Skills Chat)The in-editor AI agent you specialize with skills to design, storyboard, and generate on your canvas.
- Sign-in, Account & Public PagesSign in or create an account, accept an org invite, and find Framo's public pages.
- Canvas Extras — Notes & Screen-shareAdd sticky notes and a live screen-share capture as inputs on the infinite canvas.
- Chat Bar — Generating ImagesWrite prompts, pick models, and add references, LoRAs, looks, and aspect ratios in the floating chat bar.
- 2D Canvas Editor (Pixi)The Pixi pixel/vector editor — layers, painting, pen/shape vector tools, selections, and warp.
- 3D Scene EditorImport models, add primitives and lights, texture, frame a camera, and render the scene.
- Image AI ToolsOne-click AI operations on an image node — upscale, remove background, reframe, new angles, image-to-3D.
- Training & Using LoRAsTrain a custom AI model on your own images, then call it in generations with a trigger word.
- Framo Knowledge MCP ServerConnect an AI agent to Framo's knowledge base over the Model Context Protocol.
- Projects, Media Library & StockCreate and open projects, manage your image/video library, pull in stock photos, and move media to the canvas.
- Sharing, Public Viewer & DropsShare a project with a public link, inspect layers in the viewer, clone it, and feature it on Drops.
- Sketch to Image, Copy Design, Vectorize & EnhanceTurn sketches into images, copy a design, trace line art to vectors, and refine or reverse prompts.
- Generating VideosAnimate a still (image-to-video) or compose from references (reference-to-video) via the assistant, with per-clip confirmation.
- Workflows (Automation)Build reusable automation templates from a group of canvas nodes and replay them on any project.
Playbooks
Step-by-step walkthroughs that chain features together end to end.
- Character / design lock → multi-scene video seriesLock a subject with an agent skill, carry the reference forward, and render a multi-scene series.
- Image → Video (I2V) and Reference-to-VideoAnimate a selected image into a clip, or compose video from multiple references.
- Import a 3D model → material → lighting → render → saveBring a model into the scene, set materials and lighting, frame it, render, and export.
- Matrix → Crop — variations from one generationGenerate a grid of variations in a single image, then crop the keepers out as separate nodes.
- Prompt polish → Generate → Batch variantsPolish a prompt with AI, set a batch count, and sweep a parameter across variants.
- Sign up → create org → start generatingCreate an account, set up an organization, and run your first generation.
- Sketch → Image → 3D → ShareDraw a sketch, turn it into an image, generate a 3D mesh, then share the project.
- Stock image → canvas → render → organizeFind a stock photo, place it, refine it with AI tools, then group and arrange.
- Train a LoRA → mention trigger word → compose multiple LoRAsTrain a LoRA, invoke its trigger word in a prompt, and stack several LoRAs in one generation.