Gemini API Quota Exceeded

Your API calls are returning HTTP 429 with RESOURCE_EXHAUSTED. This is a rate limit, not a payment problem.

Immediate fix

1. Stop retrying in a tight loop. Implement exponential backoff: 1s, 2s, 4s, 8s, with random jitter and a retry cap.
2. Honour the retry-after header when the response includes one.
3. Serialise concurrent requests so you stay under requests-per-minute.

Raise the ceiling

• Enable billing on the Google Cloud project — free-tier limits are deliberately low.
• Request a quota increase for the specific model in the Cloud console. Limits are per model and per project.
• Split unrelated workloads across separate projects so a batch job cannot starve production traffic.

Send less

• Trim conversation history to the turns that still matter.
• Cache responses for repeated identical prompts.
• Use a Flash model for high-volume, low-complexity work — higher limits, lower cost.