Gemini CLI
Gemini CLI and a Gemini model ID are separate integration choices. Gemini CLI uses Google's native Gemini API protocol. A Gemini model exposed through LivRouter's OpenAI-compatible API can be used by an OpenAI-compatible client without proving native Gemini CLI compatibility.
Check protocol support first
https://api.livrouter.com/v1 for OpenAI-compatible requests. A native Gemini endpoint has not been verified here. Confirm an account-supported native Gemini route before configuring Gemini CLI; setting its base URL to the OpenAI-compatible URL does not convert the protocol.The official client supports GOOGLE_GEMINI_BASE_URL when using Gemini API-key authentication. That setting requires a server speaking the native Gemini protocol. Google sign-in and Vertex AI are distinct authentication modes.
Use Gemini models through LivRouter
- Find an accessible Gemini model ID in the Gemini catalog.
- Choose an OpenAI-compatible client such as Cline or OpenCode.
- Set its base URL to
https://api.livrouter.com/v1and use your LivRouter key. - Send a short prompt and verify the model in Usage logs.
If native Gemini access is available
Ask support for the native route, authentication requirements, available model IDs, and tool support. Apply those values using the official Gemini CLI configuration guide, then verify both a text request and a tool request. Keep native and OpenAI-compatible configurations separate.