AnyModelUse AnyModel in your tools
One endpoint, one key, every model — wired into the coding tools you already use. Pick yours for a step-by-step setup.
These are the editors, terminals and coding agents that AnyModel plugs into. Each one talks to a standard OpenAI- (or Anthropic-) compatible endpoint, so once it's pointed at AnyModel you can run GPT, Claude, Gemini and more from inside the tool you already use.
Setup falls into two camps. Tools like Claude Code, Codex CLI and OpenCode have a one-line installer that writes the config for you; the rest — Cursor, Windsurf, Zed, Cline, Aider and Continue — accept a custom base URL and key (BYOK), so you paste AnyModel's endpoint and your key into their provider settings.
Switching models is usually a one-word change to a model id, and everything runs off a single pay-per-token balance — no Claude Max, ChatGPT Plus or Gemini subscription required. Pick your tool below for a step-by-step guide.
Run Claude Code on any model through one endpoint — no Claude Max subscription required.
Connect the OpenAI Codex CLI to any model with one command.
Use any model in Cursor with a custom OpenAI-compatible endpoint.
Add any model to Windsurf Cascade via an OpenAI-compatible provider.
Wire any model into Zed's assistant with an OpenAI-compatible endpoint.
Set up OpenCode with one command and reach every model.
Use Gemini 3 (and any other model) through one endpoint.
Point Cline at any model via an OpenAI-compatible provider.
Run Aider on any model through one OpenAI-compatible endpoint.
Add any model to Continue.dev with an OpenAI-compatible provider.