Connect any AI agent to one cheap endpoint

NousResearch’s personal agent — native custom-endpoint wizard.

Set up Hermes →

Open-source personal agent with native custom-provider config.

Set up OpenClaw →

OpenAI’s terminal coding agent, pointed at our models instead of theirs.

Set up OpenAI Codex →

DeepSeek’s open-source coding agent — the Claude Code idea, running on your key.

Set up DeepSeek Harness →

Autonomous coding agent (OpenAI-Compatible via LiteLLM).

Set up OpenHands →

VS Code / JetBrains coding agent — OpenAI-Compatible provider.

Set up Kilo Code →

Override the OpenAI Base URL — bring your own key and model.

Set up Cursor →

VS Code agent — “OpenAI Compatible” provider, three fields.

Set up Cline →

The most active Cline fork — same OpenAI-Compatible provider.

Set up Roo Code →

CLI pair-programmer — two env vars and a model flag.

Set up Aider →

VS Code / JetBrains chat + edit — one models block in config.

Set up Continue →

An openai_compatible block in settings.json — max_tokens is required.

Set up Zed →

A provider block in opencode.json, on the AI SDK openai-compatible package.

Set up opencode →

One OpenAI credential with a Base URL override powers every AI node.

Set up n8n →

Add an OpenAI-API-compatible model provider in settings.

Generate Dify config →

ChatOpenAI node with a custom base path.

Generate Flowise config →

Self-hosted ChatGPT-style UI — one OpenAI API connection.

Set up Open WebUI →

Custom Chat Completion source — endpoint, key, model id.

Set up SillyTavern →

One custom endpoint block in librechat.yaml.

Set up LibreChat →

Desktop chat app — add a remote OpenAI-compatible provider.

Generate Jan config →

Self-hosted chat UI with a custom OpenAI provider slot.

Generate LobeChat config →

Desktop client — add a provider with a base URL and key.

Generate Cherry Studio config →

Pick the Generic OpenAI LLM provider and paste the endpoint.

Generate AnythingLLM config →

ChatOpenAI takes a base_url — the rest of your chain is untouched.

Set up LangChain →

OpenAILike, with is_chat_model=True — the default is False and it will bite you.

Set up LlamaIndex →

createOpenAICompatible — a provider in four lines.

Set up Vercel AI SDK →

One openai/ prefix and an api_base — as a library or as a proxy.

Set up LiteLLM →

OpenAIChatModel with an OpenAIProvider — typed agents, one base_url.

Set up PydanticAI →

LLM(custom_openai=True, base_url=...) — one object, shared by every agent.

Set up CrewAI →

Works against any OpenAI-compatible endpoint — after two settings most people miss.

Set up OpenAI Agents SDK →