Skip to main content
Version: dev

OrcaRouter

OrcaRouter provides 150+ AI models including OpenAI, Anthropic, Gemini, DeepSeek and Qwen behind a single OpenAI-compatible endpoint and API key.​

This section describes how to use the OrcaRouter provider with DB-GPT.​

  1. Sign up at OrcaRouter and generate an API key.
  2. Set the environment variable ORCAROUTER_API_KEY with your key.
  3. Use the configs/dbgpt-proxy-orcarouter.toml configuration when starting DB-GPT.

You can look up models at https://www.orcarouter.ai/models​

Or you can use docker/base/Dockerfile to run DB-GPT with OrcaRouter:​

# Expose the port for the web server, if you want to run it directly from the Dockerfile
EXPOSE 5670

# Set the environment variable for the OrcaRouter API key
ENV ORCAROUTER_API_KEY="***"

# Just uncomment the following line in the `Dockerfile` to use OrcaRouter:
CMD ["dbgpt", "start", "webserver", "--config", "configs/dbgpt-proxy-orcarouter.toml"]