Gemini API Quota Exceeded Error: Fix
What a 429 RESOURCE_EXHAUSTED actually means and the three ways to get past it.
Updated August 13, 2026
A Gemini API call that fails with HTTP 429 and RESOURCE_EXHAUSTED means you crossed a rate limit, not that you ran out of money. There are three separate limits and they are enforced independently.
The three limits
- Requests per minute (RPM) — the one most projects hit first.
- Tokens per minute (TPM) — hit by a small number of very large prompts.
- Requests per day (RPD) — hit by batch jobs on the free tier.
Each limit is per model and per project. Moving traffic from Flash to Pro does not share the budget; it uses a different one.
Fix 1: Exponential backoff
Retry after 1s, 2s, 4s, 8s with jitter, and cap the number of attempts. Retrying immediately in a loop keeps you permanently rate limited and is the reason many outages persist far longer than they should.
Fix 2: Move off the free tier
Free-tier limits are low by design. Enabling billing on the Google Cloud project raises them substantially and usually removes the problem entirely for small applications.
Fix 3: Reduce what you send
Trim system instructions, strip redundant history, and batch small requests. For high-volume classification work, a smaller Flash model has both higher limits and lower latency.
Related help
Related help
Independent resource. Not affiliated with, endorsed by, or sponsored by Google.