Integrations
Explore tools and integrations powered by the Chutes decentralized AI inference network, from coding agents to automation and chat surfaces.
One endpoint for every tool
Chutes serves an OpenAI-compatible API, so tools that speak the OpenAI API connect with the same three settings. Read model IDs, prices, and context lengths from the live catalog instead of copying them, and filter it on confidential_compute to pick models that run in a trusted execution environment.
- Base URL
- https://llm.chutes.ai/v1
- Auth header
- Authorization: Bearer cpk_...
- Model IDs
- GET https://llm.chutes.ai/v1/models
Need a tool that isn't listed? Start with the OpenAI-compatible quickstarts, browse the Agent Hub, or read about private inference.
Built on Chutes
Apps made by Chutes that run on Chutes inference, and the sign-in they use.
Chutes Chat
Chat app
Chat with open-source AI models on Chutes in the browser: switch models mid-conversation, search the web, generate and edit images, and attach files.
fictio
Character chat and roleplay
fictio is an AI character chat and roleplay platform built by Chutes. Accounts use Sign in with Chutes.
Sign in with Chutes
OAuth 2.0
Let users sign into your own app with their Chutes account. API calls run on the user's behalf, and usage is billed to their Chutes account, not yours.
Coding agents and editors
Run open-source models from your editor or terminal agent.
Claude Code
BetaAI coding agents
Install the Chutes plugin for Claude Code or Cowork: skills and an MCP server for account setup, keychain-stored cpk_ keys, model discovery, inference, routing, usage, and TEE evidence.
Cursor
BetaEditors
Point Cursor’s OpenAI provider at https://llm.chutes.ai/v1 to run open-source models in chat and agent mode, or expose Chutes tools to Cursor over MCP.
Cline
BetaEditors
Wire Chutes into Cline through its global MCP settings file. The agent toolkit generates the snippet and keeps CHUTES_API_KEY in the environment.
Aider
BetaTerminal agents
Point Aider at Chutes over OpenAI-compatible HTTP, with Bearer auth and a live model ID from /v1/models.
Codex
Terminal agents
Guide only: use Chutes wherever the Codex runtime accepts an OpenAI-compatible base URL, and fall back to a sidecar, SDK script, or MCP bridge for locked runtimes.
MCP clients
BetaDeveloper tools
The Chutes MCP server exposes model discovery, quota, usage, and API key read tools to Cursor, Cline, Claude Desktop, and other MCP clients. Write tools stay labelled beta.
Agent frameworks and automation
Route framework and workflow model calls through Chutes.
Hermes
Agent frameworks
Run Nous Research Hermes agents with Chutes as a custom OpenAI-compatible provider, using live model discovery and explicit routing caveats.
OpenClaw
BetaAgent gateways
OpenClaw is a multi-channel AI gateway for agents. Put Chutes behind its channel agents as an OpenAI-compatible provider; the installer at chutes.ai/openclaw sets it up and starts a local gateway in one command.
LangChain
Frameworks
Call Chutes from LangChain’s ChatOpenAI client by changing base_url, api_key, and the model to a live catalog ID.
LiteLLM
Frameworks
Send LiteLLM chat requests to Chutes with api_base set to https://llm.chutes.ai/v1 and a Bearer-compatible key.
Vercel AI SDK
Frameworks
Use an OpenAI-compatible provider configured with baseURL, apiKey, and a live model ID from the Chutes catalog.
n8n
Automation
Community nodes and a local installer connect Chutes inference to n8n workflows, including Sign in with Chutes.
Any OpenAI-compatible agent
System prompt
For agents without a dedicated guide, the Chutes system prompt covers URLs, auth, live model discovery, routing, and TEE boundaries.
Roleplay and chat frontends
Use Chutes models in the chat and roleplay apps you already run.