Gemini 3.8 Flash TTS – API Enforces 100 RPD Despite AI Studio Showing 10K RPD

Gemini API
1

Hi Google AI Team,

I’m experiencing an unexpected daily quota limit when using the Gemini 3.8 Flash TTS API.

Project Information

  • Project ID: gen-lang-client-05003270..

  • Model: gemini-3.8-flash-tts

  • API: generateContent

Issue Description

According to the Google AI Studio rate limits dashboard, my project currently shows:

  • RPM: 25 / 1,000

  • TPM: 4.17K / 100K

  • RPD: 193 / 10,000

However, API requests are returning HTTP 429 (RESOURCE_EXHAUSTED) with the following quota violation:

{
  "quotaMetric": "generativelanguage.googleapis.com/generate_requests_per_model_per_day",
  "quotaId": "GenerateRequestsPerDayPerProjectPerModel",
  "quotaDimensions": {
    "location": "global",
    "model": "gemini-3.8-flash-tts"
  },
  "quotaValue": "100"
}

Expected vs. Actual Behavior

  • Expected: The API should allow up to 10,000 requests per day, as displayed in AI Studio.

  • Actual: The API enforces a limit of 100 requests per day, despite the dashboard showing a 10,000 RPD quota.

Questions

  1. Why is the API enforcing a 100 RPD limit when AI Studio explicitly shows 10,000 RPD?

  2. Is this a known issue or a mismatch between the displayed quota and the backend quota enforcement?

  3. Is there an additional model-specific quota that is not reflected in the dashboard?

  4. Could the Google AI team investigate the quota configuration for my project and advise how to resolve this discrepancy?

I’ve attached a screenshot of the AI Studio rate limits dashboard for reference.

Thank you for your help!

image
image2048×107 16.3 KB