Gemini 3 Free Search Grounding Quota is locked to 0 in Cloud Console (Pricing says 5,000 free/month)

Hi Community Team & Developer Advocates,

I am reporting a platform provisioning mismatch between the published Gemini API pricing documentation and the Google Cloud Quotas engine for Gemini 3 models.

The Pricing Promise

According to ai.google.dev/gemini-api/docs/pricing:

“5,000 free search requests per month (shared across all Gemini 3 and newer models), then $14 per 1,000 requests.”

The Issue in Cloud Console

In Google Cloud Console under IAM & Admin > Quotas & System Limits for metric generativelanguage.googleapis.com/generate_content_search_request_free:

  • Legacy dimensions (gemini-2.0, gemini-2.5) receive an active daily allowance (1,500/day).
  • The model : gemini-3 dimension is hardcoded to 0.
  • When clicking “Edit Quota” to request an adjustment, the UI displays a hard validation error:
    "The new quota value must be between 0 and 0."

Impact on Developers

Because gemini-3 has 0 free quota provisioned on the backend:

  1. SKU 7166-DCCD-7D46 (“Generate content search query gemini 3 free”) is completely bypassed.
  2. Every single search grounding call is routed directly to paid SKU E662-8171-51CB (“Generate content search query gemini 3 paid one”) at $14/1,000 requests from the very first call.
  3. This recently drained an entire prepay balance unexpectedly.

Could a Google Developer Advocate or API Team Lead look into this backend quota definition for model : gemini-3?

  • Project ID: gen-lang-client-0277997847
  • Screenshots showing the Quotas table, the “0 to 0” lock, and the SKU cost breakdown are attached below.

Thank you!

Hello @Vivek_Ramoliya,

  • The 0 limit in Cloud Console applies only to Free Tier projects and does not affect your Paid Tier 5,000/month free search allowance (Gemini API Pricing).
  • Please check Usage (units) rather than cost in Cloud Billing Reports to view your free queries, as free SKUs always appear as 0.00 on cost charts.
  • Each search query the model runs counts toward the 5,000 monthly allowance, with model tokens billed separately (Grounding with Google Search). If anything still looks unexpected, please reach out to Cloud Billing Support so they can review your account.