Gemini API Free Tier daily quota does not reset — persistent 429 RESOURCE_EXHAUSTED

Gemini API
1

Hello Google Gemini API Support,

I am experiencing a persistent quota problem with the Gemini API used through Antigravity CLI.

My project is on the Gemini API Free Tier.

The affected models are:

  • gemini-3.8-flash
  • gemini-3.7-flash
  • gemini-3.6-flash

All three models immediately return HTTP 429 RESOURCE_EXHAUSTED.

The error explicitly states:

Quota exceeded for metric:
generativelanguage.googleapis.com/generate_content_free_tier_requests

quotaId:
GenerateRequestsPerDayPerProjectPerModel-FreeTier

quotaValue: 20

For example:

model: gemini-3.7-flash
quotaId: GenerateRequestsPerDayPerProjectPerModel-FreeTier
quotaMetric: generativelanguage.googleapis.com/generate_content_free_tier_requests
quotaValue: 20
status: RESOURCE_EXHAUSTED

The same type of error occurs with gemini-3.8-flash and gemini-3.6-flash.

According to the official Gemini API documentation, Requests Per Day (RPD) quotas reset at midnight Pacific Time.

However, after the expected daily reset, the models still immediately return the same daily quota exceeded error.

There is another issue I would like you to investigate.

A single task submitted to Antigravity appears to exhaust the 20-request daily quota extremely quickly. Please clarify whether Antigravity makes multiple internal Gemini API requests for a single user task and whether every internal request is counted separately against the RPD quota.

I checked the Google AI Studio Rate Limit page. It does not show the corresponding 20/20 RPD usage for gemini-3.8-flash, gemini-3.7-flash and gemini-3.6-flash, even though the API responses explicitly say that the daily Free Tier quota of 20 has been exhausted.

Please investigate:

  1. Why did the Free Tier daily RPD quota not reset at midnight Pacific Time?

  2. Why do gemini-3.8-flash, gemini-3.7-flash and gemini-3.6-flash immediately return quotaValue: 20 / RESOURCE_EXHAUSTED after the expected reset?

  3. Is every internal Antigravity agent/model call counted as a separate Gemini API request toward the 20 RPD quota?

  4. Why is this model-specific 20-request quota usage not visible on my Google AI Studio Rate Limit page?

  5. Is there a quota provisioning, entitlement, or backend problem affecting my project?

I can provide my Google Cloud Project Number privately to a Google employee for investigation.

I also have screenshots and the complete 429 RESOURCE_EXHAUSTED responses for all three affected models.

Please check the backend quota state for my project.

Thank you.

2

Hello @Valeriy197219 ,

If you’re receiving 429’s and have checked no rate limit was breached for the period, please add your details to this form & we’ll check it out.