Vertex AI Anthropic models return 429 RESOURCE_EXHAUSTED for weeks, quota shows only 0..0

Google Cloud Build with AI Generative AI & Foundational Models
1

Hi,
we have a persistent issue with Vertex AI Anthropic API calls for ~2 weeks.
Project:

  • Endpoint / method:
  • aiplatform.googleapis.com
  • streamRawPredict
    Models tested:
  • claude-haiku-4-5@20251001
  • claude-sonnet-4-6
    Result:
  • HTTP 429 RESOURCE_EXHAUSTED (reproducible)
  • Same behavior via official AnthropicVertex Python client and direct curl
    Important detail:
  • In Quotas, relevant metrics can only be set from 0 to 0 (cannot request increase)
    We currently cannot open Google Cloud support cases due to plan limitations.
    Could someone from Google verify if this is an entitlement/capacity lock on the project and how to unlock Anthropic model quotas for this project?
    Thanks.