# NodeTool Documentation > Local-first visual environment for building and running AI workflows. Build agents visually, deploy anywhere, privacy by design. NodeTool is an open-source, local-first visual environment for building AI workflows. Wire image, video, audio, text, and agent models on a node-based canvas. Runs on macOS, Windows, Linux (Studio), or in the browser (Cloud). AGPL-3.0. ## Getting Started - [Introduction](https://docs.nodetool.ai/): Overview of NodeTool — Studio vs Cloud, what you can build. - [Why NodeTool](https://docs.nodetool.ai/comparisons): How NodeTool compares to ComfyUI, n8n, LangChain, etc. - [Installation](https://docs.nodetool.ai/installation): Install NodeTool Studio on macOS, Windows, or Linux. - [Quick Start](https://docs.nodetool.ai/getting-started): Install, run a template, edit and iterate. - [Agents and Automation](https://docs.nodetool.ai/agents/): Install, discover node schemas, validate and run workflows programmatically. ## Core Concepts - [Key Concepts](https://docs.nodetool.ai/key-concepts): Canvas, nodes, edges, workflows, agents, providers. - [Workflows](https://docs.nodetool.ai/workflow-editor): Building, running, and managing workflow graphs. - [Nodes](https://docs.nodetool.ai/nodes/): Catalog of available nodes by category. - [Node catalog](https://docs.nodetool.ai/nodes/catalog.json): Machine-readable node types, inputs, outputs, and documentation URLs. - [Agents](https://docs.nodetool.ai/global-chat-agents): Planning agents, tools, multi-step execution. - [Models](https://docs.nodetool.ai/models-and-providers): Local and cloud model providers. - [Glossary](https://docs.nodetool.ai/glossary): Terminology used across NodeTool. ## Workflow Editor - [User Interface](https://docs.nodetool.ai/user-interface): UI tour and shortcuts. - [Workflow Editor](https://docs.nodetool.ai/workflow-editor): Canvas, ports, edges, execution. - [Editor Panels](https://docs.nodetool.ai/editor-panels): Inspector, assets, models, chat panels. - [Chain Editor](https://docs.nodetool.ai/chain-editor): Linear chain authoring mode. - [Graph View](https://docs.nodetool.ai/workflow-graph-view): Read-only graph navigation. - [Debugging](https://docs.nodetool.ai/workflow-debugging): Inspect runs, errors, logs. - [Sketch Editor](https://docs.nodetool.ai/sketch-editor): Mask, paint, layer, and generate images inline. - [Templates Gallery](https://docs.nodetool.ai/templates-gallery): Browse and open starter workflows. - [App Builder](https://docs.nodetool.ai/app-builder): Design Mini App interfaces for workflows. ## Chat & Agents - [Chat](https://docs.nodetool.ai/global-chat): Provider-agnostic chat interface. - [Chat & Agents](https://docs.nodetool.ai/global-chat-agents): Use agents from chat. - [Agent Config](https://docs.nodetool.ai/agent-config-schema): YAML schema for agent definitions. - [Agent Memory](https://docs.nodetool.ai/agent-memory): How agents persist context. ## Models & Providers - [Overview](https://docs.nodetool.ai/models-and-providers): All supported providers. - [Local vs Cloud](https://docs.nodetool.ai/models): When to use which. - [Provider Reference](https://docs.nodetool.ai/providers): Per-provider configuration. - [HuggingFace](https://docs.nodetool.ai/huggingface): Models from the Hub. - [Models Manager](https://docs.nodetool.ai/models-manager): Download and manage local models. ## Examples - [Cookbook](https://docs.nodetool.ai/cookbook): End-to-end patterns and recipes. - [Core Patterns](https://docs.nodetool.ai/cookbook/core-concepts): Foundational workflow patterns. - [Pipeline Patterns](https://docs.nodetool.ai/cookbook/patterns): Composition patterns. - [Workflow Examples](https://docs.nodetool.ai/workflows/): Ready-to-run sample workflows. ## Configuration - [Configuration](https://docs.nodetool.ai/configuration): Settings and environment variables. - [Storage](https://docs.nodetool.ai/storage): Local and S3 storage backends. - [Vector Storage](https://docs.nodetool.ai/vector-storage): SQLite-vec (default), Pinecone, Supabase pgvector. - [Indexing](https://docs.nodetool.ai/indexing): Document ingestion and RAG pipelines. - [Collections](https://docs.nodetool.ai/collections): Vector collections for RAG. - [Asset Management](https://docs.nodetool.ai/asset-management): Files, images, datasets. - [Authentication](https://docs.nodetool.ai/authentication): Auth setup for hosted deployments. - [Security Hardening](https://docs.nodetool.ai/security-hardening): Production security guidance. ## Deployment - [Deployment Guide](https://docs.nodetool.ai/deployment): Where and how to deploy. - [End-to-End Guide](https://docs.nodetool.ai/deployment-e2e-guide): Full production walkthrough. - [Self-Hosted](https://docs.nodetool.ai/self-hosted-deployment): Run on your own infra with Docker. - [Docker Resources](https://docs.nodetool.ai/docker-resource-management): GPU, memory, volumes. - [Worker Deployment](https://docs.nodetool.ai/worker-deployment): Rent GPU workers (RunPod, Vast), attach, tear down. - [Supabase](https://docs.nodetool.ai/supabase-deployment): Postgres + auth backend. ## CLI - [CLI Overview](https://docs.nodetool.ai/cli): `nodetool` command reference. - [Chat CLI](https://docs.nodetool.ai/chat-cli): Interactive chat with agent mode. - [Agent CLI](https://docs.nodetool.ai/agent-cli): Run agent configs from the terminal. - [Chat Server](https://docs.nodetool.ai/chat-server): Serve the OpenAI-compatible chat API via `nodetool serve`. - [API Server](https://docs.nodetool.ai/api-server): REST + WebSocket API server. ## API Reference - [API Overview](https://docs.nodetool.ai/api-reference): REST API conventions. - [Workflow API](https://docs.nodetool.ai/workflow-api): Create, run, manage workflows. - [WebSocket API](https://docs.nodetool.ai/websocket-api): Streaming run protocol. - [Chat API](https://docs.nodetool.ai/chat-api): OpenAI-compatible chat endpoint. ## Developers - [Developer Guide](https://docs.nodetool.ai/developer/): Build and extend NodeTool. - [Architecture](https://docs.nodetool.ai/architecture): Backend packages, kernel, runtime, providers. - [Packages](https://docs.nodetool.ai/packages): NPM workspace package map. - [Python Bridge Protocol](https://docs.nodetool.ai/python-bridge-protocol): Stdio msgpack bridge to Python nodes. - [Execution Strategies](https://docs.nodetool.ai/execution-strategies): Actor-model workflow execution and code-runner sandboxes. - [Custom Nodes](https://docs.nodetool.ai/developer/custom-nodes-guide): Implement BaseNode subclasses. - [Node Patterns](https://docs.nodetool.ai/developer/node-patterns): process / genProcess / streaming. - [Node Reference](https://docs.nodetool.ai/developer/node-reference): @prop, types, decorators. - [TypeScript DSL](https://docs.nodetool.ai/developer/ts-dsl-guide): Define workflows in TS. - [Node Packs](https://docs.nodetool.ai/node-packs): Distribute nodes as installable packs. ## Apps - [Desktop App](https://docs.nodetool.ai/desktop-app): NodeTool Studio (Electron). - [Desktop App Views](https://docs.nodetool.ai/electron-views): Windows and panels. - [Mobile App](https://docs.nodetool.ai/mobile-app): React Native client. ## Help - [Troubleshooting](https://docs.nodetool.ai/troubleshooting): Common errors and fixes. - [Tips & Tricks](https://docs.nodetool.ai/tips-and-tricks): Productivity shortcuts.