Low rate limits batch embeddings Tier 3

Gemini API
,
1

Hello,

I want to embed a very large dataset (>60 billion tokens) using gemini-embeddings-001 using the batch embeddings to save costs.

  1. I’m on Tier 3, but the rate limit remains at 10,000,000 tokens enqueued. This seems low compared to other models (e.g. Gemini 3 Pro Preview is at 1,000,000,000) and to the . I have filled in the form to receive higher rate limits but get no response. How do I receive higher rate limits?
  2. Even though I have just under 10 million tokens enqueued, I don’t see my usage in the rate limits in the Quota overview on Google Cloud. Is this a known bug?
1 Like