Skip to main content
Anyone aggregates 40+ AI providers. You only need one API key to call all models — no need to register separately with OpenAI, Anthropic, Google, etc. Plus, Anyone automatically handles format conversion, so you can use the OpenAI SDK to call Claude and Gemini.
Fully supported. Just set base_url and api_key:
from openai import OpenAI

client = OpenAI(
    base_url="https://api.anyone.ai/v1",
    api_key="YOUR_TOKEN",
)
All features supported by the OpenAI SDK (chat completions, streaming, function calling, etc.) work out of the box.
Yes. Send a standard /v1/chat/completions request with the model set to a Claude model name like claude-sonnet-4-6. Anyone automatically converts the request to Claude Messages format and converts the response back to OpenAI format.The reverse also works — Anyone’s /v1/messages endpoint supports the native Claude format.
Yes. Set "stream": true in your request:
{
  "model": "gpt-4o",
  "stream": true,
  "messages": [{"role": "user", "content": "Hello"}]
}
Responses are returned in real time using SSE (Server-Sent Events) format.
Yes. Pass the tools parameter in your request, exactly as you would with OpenAI. Anyone automatically converts tool calling to the format supported by the target model.
Click Forgot password on the login page and follow the instructions to reset it.
Immediately log in to the dashboard and delete the token, then create a new one. A deleted token is invalidated instantly.
Call the /v1/models endpoint for the full list:
curl https://api.anyone.ai/v1/models \
  -H "Authorization: Bearer YOUR_TOKEN"
You can also check the models page in the dashboard.
Any tool compatible with the OpenAI API works, including:
  • Cherry Studio — set the base URL to https://api.anyone.ai/v1
  • LangChain / LlamaIndex — change the base URL in the configuration
  • Vercel AI SDK — use the OpenAI provider and change the base URL
  • Cursor / Continue — set a custom API endpoint
  • And any other tool that supports a custom OpenAI endpoint
See the troubleshooting page for common error codes and solutions.