Batch API returns 429 RESOURCE_EXHAUSTED despite being well under quota limits

[poliha]

Environment

• SDK: @google/genai v1.37.0
• Node.js: 22.x
• Model: gemini-2.5-flash-lite
• Account: Paid billing enabled

Description

Consistently receiving 429 RESOURCE_EXHAUSTED errors when using the Batch API,
despite usage being well below documented quota limits.

Error Message

status: 429
message: "You exceeded your current quota, please check plan and billing
details. For more information on this error, head to:
https://ai.google.dev/gemini-api/docs/rate-limits"

Details

• Batch size: 1000 jobs (~5M tokens total)
• Only 1 concurrent batch in flight
• AI Studio dashboard shows no quota exceeded
• Adding delays between batch creations (10+ seconds) doesn't help
• First batch succeeds, subsequent batches fail with 429

Related Issue

This mirrors the exact behavior reported in the Python SDK:
googleapis/python-genai#1901

A comment there suggests Vertex AI (vertexai: true) has higher quota limits
than the Generative Language API. Can you confirm if this is expected behavior
or a bug?

Reproduction

1. Create batch with ~1000 requests using genAI.batches.create()
2. Wait for completion
3. Create another batch
4. Second batch fails with 429

[shmishra99]

@poliha , I have added you to a private Gist where you can share your Google Cloud project ID, so we can check this issue.

[Varun-S10]

Hi @poliha, We have passed this issue to our internal team for further investigation. Thank you for your patience.

[idanrk]

@Varun-S10 Hi we are facing the same problem.. What is the status on this issue?

[cofirazak]

same for me

[wejeel]

Same here. Is there a fix for this on the way and will submitted jobs eventually complete ?

[Varun-S10]

Hi @poliha, @idanrk, @cofirazak, @wejeel, Could you please let me know whether you are facing this issue using the Gemini API or Vertex AI?

[wejeel]

Yes, I am still facing the same issue and batch API is taking more than 24 hours. I thought the SLA was 24 hours, it’s well over 72 hours for a batch write job of 15k requests.

…

On Tue 10 Mar 2026 at 07:23, Varun-S10 ***@***.***> wrote: *Varun-S10* left a comment (googleapis/js-genai#1264) <#1264 (comment)> Hi @poliha <https://github.com/poliha>, @idanrk <https://github.com/idanrk>, @cofirazak <https://github.com/cofirazak>, @wejeel <https://github.com/wejeel>, Could you please let me know whether you are facing this issue using the Gemini API or Vertex AI? — Reply to this email directly, view it on GitHub <#1264 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ADECQLUNFJZKAA7SMJ3GFMD4P67I3AVCNFSM6AAAAACSDSMY52VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHM2DAMRZGI3DAMRUGU> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[Varun-S10]

@wejeel, To help me troubleshoot, could you specify which platform you are facing this issue on-the Gemini API or Vertex AI?