The open creative AI workspace
One node-based canvas for making images, video, sound, and stories with AI. Wire up Flux, Qwen, Wan, Seedance, Sora, Veo, Kling, ElevenLabs, MusicGen, and the major LLMs side by side — bring your own keys, or run locally. Open source, AGPL-3.0.
Both editions are open source under AGPL-3.0 and built from the same code. Workflows are portable between them.
Runs on macOS, Windows, and Linux. Local inference via Ollama, MLX, and GGUF on your hardware. Works offline. Files, prompts, and outputs stay on disk. Bring your keys for cloud providers when you want them.
For: local inference, offline work, GPU owners, privacy.
Download Studio →Hosted, no install. Same canvas, same nodes. Bring your keys for every cloud provider — OpenAI, Anthropic, Gemini, Replicate, FAL, ElevenLabs, HuggingFace. Does not run local models.
For: no setup, working across devices, no GPU.
Open Cloud →No credit markup. Cloud is managed hosting of the code in this repo. You can self-host the same Docker images and CLI any time. You pay providers directly.
Posters, characters, scenes. Flux, Qwen, Wan, Seedance, Sora, Veo, Kling on one canvas.
Movie Posters →Turn a prompt into a storyboard, narrate it, animate it, score it.
Story to Video →Music, sound design, narration. ElevenLabs, MusicGen, Whisper wired into the graph.
Image to Audio Story →Multi-step agents that plan, call tools, and drive your creative pipelines.
Realtime Agent →More patterns — pipelines, data, RAG, email — in the Cookbook.
