What happened?
API requests made through the Gemini CLI can hang indefinitely (or up to the default Node.js/undici timeout of 5 minutes) when the backend or local proxy becomes unresponsive. This results in a poor user experience where the CLI appears frozen.
What did you expect to happen?
The CLI should have a reasonable, explicit client-side timeout (e.g., 60 seconds) for API calls. If this timeout is exceeded, the request should be aborted, and the error should be treated as a retryable ETIMEDOUT error so that the built-in resilience logic (exponential backoff) can attempt the request again.
Client information
Linux
Login information
No response
Anything else we need to know?
No response
What happened?
API requests made through the Gemini CLI can hang indefinitely (or up to the default Node.js/undici timeout of 5 minutes) when the backend or local proxy becomes unresponsive. This results in a poor user experience where the CLI appears frozen.
What did you expect to happen?
The CLI should have a reasonable, explicit client-side timeout (e.g., 60 seconds) for API calls. If this timeout is exceeded, the request should be aborted, and the error should be treated as a retryable ETIMEDOUT error so that the built-in resilience logic (exponential backoff) can attempt the request again.
Client information
Linux
Login information
No response
Anything else we need to know?
No response