Repository navigation
gemini-2.5-pro is [API Error: got status: 429 Too Many Requests. #1502
Description
✕ [API Error: got status: 429 Too Many Requests. {"error":{"message":"{\n \"error\": {\n
\"code\": 429,\n \"message\": \"Resource has been exhausted (e.g. check quota).\",\n
\"status\": \"RESOURCE_EXHAUSTED\"\n }\n}\n","code":429,"status":"Too Many Requests"}}]
why?
Activity
arkavo-com commentedon Jun 25, 2025
same here. I have gotten 57 pro and 40 flash requests. I am using my GEMINI_API_KEY
Tombar commentedon Jun 25, 2025
same here
AFMiziara commentedon Jun 25, 2025
Same here
fhinkel commentedon Jun 25, 2025
We're working on it! The response has been incredible, so if you don't want to wait you can also get started with an API key from: https://aistudio.google.com/app/apikey Huge apologies!
PostboxRetinal commentedon Jun 25, 2025
Just created an API from it and even after that it keeps saying resource exhausted, just tried with 2 accounts. Worked by changing auth method to google account directly
bbhoss commentedon Jun 25, 2025
Not trying to pile on but nobody has mentioned that this appears to be affecting paid API keys, not just free accounts. It seems like the free accounts are working more reliably while the paid keys are getting blocked. Sometimes I get a single response but this it quickly fails with a paid key. Quotas in cloud console show quota being utilized, but not near any limits and I never get any meaningful responses, much less changes to my codebase.
ZackMattor commentedon Jun 25, 2025
This is affecting us with tier 2 (paid) API keys... Excited to try this when it settles down a bit :)
Thanks for all the hard work @fhinkel! Let us know if there's anything we can do to help.
grayodesa commentedon Jun 25, 2025
Same here — 2 paid API keys, I can see in AI Studio stats some requests and token usage but never got any valid response, only 429 errors.
Zig1375 commentedon Jun 25, 2025
same here
ThroughIris commentedon Jun 25, 2025
Same here, I got a 429 error “too many requests” almost immediately, far from the stated 1000 requests, it edited 2 files then tried to switch to Gemini flash but said “rate limiting detected”. Even if I restart it gets stuck is this loop on the first request and freezes up.
tybalex commentedon Jun 25, 2025
same here, tier 2 user.
199 remaining items
vaelen commentedon Sep 28, 2025
Same here. I got in 40 requests this time before it shut me off.
theraghavjuneja commentedon Oct 7, 2025
Did anyone find out how to solve this?
Even with paid account it comes, sometimes my users increase which led to burst in requests . Yesterday there were 86 requests out of which 21 failed
mystilleef commentedon Oct 7, 2025
The errors are back with a vengeance. I now get this error after every request. I refuse to switch to my API key after spending close to $100 last month for a rather sub-optimal experience (half of my tokens were consumed by errors, loops, mistakes, etc) when compared to Claude.
hadifarnoud commentedon Oct 8, 2025
my main issue is there is no clear way of knowing if I have a rate limit or usage limit. even in AI Studio, it only shows the usage. it does not say how much left until my daily limit hourly limit etc.
lennartkoester commentedon Oct 9, 2025
On a paid Gemini Code Assist Standard subscription. I get this error message about every other query:
✕ [API Error: [{ "error": { "code": 429, "message": "Resource exhausted. Please try again later. Please refer to https://cloud.google.com/vertex-ai/generative-ai/docs/error-code-429 for more details.", "errors": [ { "message": "Resource exhausted. Please try again later. Please refer to https://cloud.google.com/vertex-ai/generative-ai/docs/error-code-429 for more details.", "domain": "global", "reason": "rateLimitExceeded" } ], "status": "RESOURCE_EXHAUSTED" } } ]] Possible quota limitations in place or slow response times detected. Switching to the gemini-2.5-flash model for the rest of this session.Funny thing is: gemini-cli doesn't switch to flash - it stays in pro and I can just repeat the last message. It's really REALLY annoying though
jains99 commentedon Oct 10, 2025
yeah same issues
aronsuarez commentedon Oct 10, 2025
same, start useing gemini cli today, first command was /init stopt workinh with 429 after few seconds. I'm on a pro plan via google workspace
ceoAppsknight commentedon Oct 10, 2025
Same happening with me. Just bought paid plan and got 429 with just /init and just used
Model: gemini-2.5-pro
Reqs: 7
Input: 52,280
Output: 319
shanevcantwell commentedon Oct 11, 2025
It seemed like when I upgraded from PRO to ULTRA, my API quotas were cut to literally like ~5% total of PRO's limits. However, I don't know how much that affects VS Code Gemini Code Assist, since I also think there's a fundamental disconnect with having to pay for 2 subscriptions to access basically the same compute.
At any rate, I get 429 errors with a Gemini Code Assist paid starting with ~1,200 turns not 10,000 as claimed in the Cloud API docs to be the quota target I should expect. This has been consistent for weeks, but seeming oddly to have gotten worse with my plan upgrades, coincidentally if not causally. I'm not sure what is actually a useful data point when there was such a close correlation in timing.
niicholai commentedon Oct 24, 2025
Just chiming in to state this is happening regularly to myself and still ongoing. Doesn't matter what account I use as I have a AI Studio Pro and a Free account. Both keep giving the 429 error either every other request or mid-request if it's doing something that takes more than 10 seconds.
theraghavjuneja commentedon Oct 26, 2025
https://ai.google.dev/gemini-api/docs/rate-limits
Each model has rate limits. Request them to increase
shanevcantwell commentedon Nov 9, 2025
I think my issue had to do with changing plans but using the same API key. Issuing a new key seems to have fixed it.
wolfgang555 commentedon Dec 7, 2025
seems not, I changed a new one, still did not work
heinsenberg82 commentedon May 30, 2026
The same issue is happening to me. Google AI Studio shows that I’m well, well below any quota or rate limit.
This problem has been reported by hundreds of users for months, and yet no solution has been provided. At this point, it makes Google’s AI models completely unusable and unreliable.
Honestly, it’s embarrassing.
tennyson-mccalla commentedon May 31, 2026
Isn't all of this just abandonware now that they're moving to the closed source antigravity-CLI?