I’m using n8n with the Google Gemini API (Video → Analyze Video) and the model gemini-3.1-pro-preview.
I keep getting this error:
Request failed with status code 429
Quota exceeded for metric: generate_content_free_tier_requests, limit: 0
The service is receiving too many requests from you.
I’m only sending a single request.
Context:
Free tier account
Billing not enabled (i’m available to pay)
New API key
Model: gemini-3.1-pro-preview
My questions:
Does Gemini 3.1 Pro require billing enabled?
Does limit: 0 mean the model is not available for free tier?
@antina Free tier does not have access to Gemini 3 Pro. Free tier has access to Gemini 2.5 Pro and 3 / 2.5 Flash / Lite). In order to use 3.1 Pro, enable billing and use a paid API key (start paying).
Please note that the free tier has very low limits too. Therefore, soon after this, you will face the same error on the other models too.
Your usage of the Gemini API can be monitored in Google AI Studio under Dashboard > Usage.
Hi @antina Personally i was also facing this problem, then i revoked my current credentials and created a new credentials and that seems to be working really good for me, consider getting a new API from AI Studio: (that would work):
Yeah limit: 0 means exactly what it says — that model has zero free tier quota. Gemini 3.1 Pro is paid only, so you need to enable billing in your Google Cloud project before it’ll work. If you just want to test things out without paying yet, switch to gemini-2.5-flash which does have a free tier.
The error message ‘limit: 0’ is the key clue here — it means the free tier quota for that specific model is literally set to zero requests. This isn’t a rate limit in the usual sense; it means Gemini 3.1 Pro is not available on the free tier at all. You need billing enabled to access it.
To fix this: go to Google Cloud Console, enable billing on your project, then generate a new API key (or use the existing one — the key itself doesn’t need to change, the billing enablement unlocks access). Once billing is active, you’ll move from the free tier quotas to the paid tier, which gives you actual per-request usage at whatever the current pricing is.
A few things to be aware of once you’re on a paid key: Gemini 3.1 Pro is a heavy model and the cost per token is meaningful, especially for video analysis which sends a lot of data. If cost is a concern, consider whether ‘gemini-2.5-flash’ or ‘gemini-2.0-flash’ would handle your use case — they’re significantly cheaper and faster, and for most document/video summarization tasks they perform very close to the Pro models. You can always test both in AI Studio first before committing.
One more tip: when working with video analysis in n8n specifically, make sure the file URI approach is correct. If you’re sending a YouTube URL or a Google Drive link, it goes through the Files API differently than a direct upload. If you’re still getting errors after enabling billing, check what the actual request payload looks like in the n8n execution log to confirm the file is being passed correctly.
