Open-Source Visual AI Workflow Builder
Connect nodes to generate content, analyze data, and automate tasks. Run models locally or via cloud APIs.
NodeTool ships in two editions, both 100% open source under AGPL-3.0 and built from the same codebase. Pick the one that fits how you want to work — workflows are portable between them.
The desktop app for macOS, Windows, and Linux. Runs Ollama, MLX, and GGUF models on your hardware. Works offline. Your data, prompts, and outputs stay on your disk. BYOK for cloud providers when you want them.
Best for: privacy, offline work, free local inference, GPU owners.
Download Studio →The hosted version. Sign in from any browser — no install, no GPU, no driver setup. Always on the latest release. Doesn't run local models, but supports BYOK for every cloud provider (OpenAI, Anthropic, Gemini, Replicate, FAL, ElevenLabs, HuggingFace, …).
Best for: zero setup, working across devices, teams, no-GPU environments.
Open Cloud →Always open source. Cloud is just our managed hosting of the code in this repo. There is no separate proprietary version — you can self-host the same Docker images, CLI, and websocket server any time.
Build multi-step agents that plan and execute tasks.
Realtime Agent →Index documents, search with AI, and answer questions from your sources.
Chat with Docs →Generate and transform media with Flux, Qwen, and other models.
Movie Posters →Transform data, extract information, and automate reports.
Workflow Patterns →NodeTool is open-source under AGPL-3.0. Join the Discord, explore the GitHub repo, and share workflows with other builders.
