400, 500 and 503 since morning

Gemini API
, , ,
1

My application started throwing 400 errors saying my input token count exceeds the maximum number of tokens allowed (131k) even though I’m a paying customer that shouldn’t be subjected to the free tier limits.

It now switched to being fully unusable with 500/503 errors.

3 Likes
2

Hi @PatrickMaia,

May I know which model/s are you using?

1 Like
3

Sorry, i forgot to mention these errors are exclusive to Gemini 2.5 Pro. I don’t run into any 429/rate limit/max tokens in the Gemini 2.5 Flash which is weird.

1 Like
4

The recent issues affecting AI studio and Gemini API have been resolved, and all systems are operational again. We apologize for the inconvenience and disruption this may have caused to your applications and workflows. Please let us know if you continue to face any issues.

1 Like
5

Not for me:

Gemini 2.5 Pro

 {"args":{},"error_type":"ClientError","message":"400 INVALID_ARGUMENT. {'error': {'code': 400, 'message': 'The input token count (860489) exceeds the maximum number of tokens allowed (131072).', 'status': 'INVALID_ARGUMENT'}}"}"}

I even tried creating a different api-key/project etc but it’s still not working and it was fine this until this morning!

Again, 2.5 Flash works, i have a paid account. What changed?

1 Like
6

I just wanted to let you know that I’m still running into the same issue. The API error is happening on a project I only created yesterday, and I’m on a paid plan. To make things more confusing, yesterday several of my Google Cloud projects disappeared out of nowhere, so I had to recreate them and get new API keys.

7

@Krish_Varnakavi1 Thank you for the update. My api key in AI studio still shows unavailable.Should I just wait or do I need to take any action? Any help is much appreciated.

1 Like
9

Yes, same issue still happening here Gemini 2.5 Pro (paid, with API key) errors 100% of the time when token count is over 131k - #7 by Steven_Waterman

1 Like
11

Any news here? More than 24 hours with our production systems facing this issue. Unacceptable for a paying customer.

12

Not for me, the error is still happening now with even paid APIs

13

Still facing this bug

14

I had the same issue for model Gemini-2.5-pro

15

Been relying on Gemini 2.5 Pro for too long without checking the competition. Just replaced everything with Grok 4 Fast Reasoning and everything is up and running again.

1 Like
16

Hi @PatrickMaia, @Marlon_Almeida @Shihela_Riannita @andrew-may @Steven_Waterman @H_i_Nguy_n1

Can you all share

  1. Model being used
  2. Your Tier
  3. Full error message
17

Hi @Krish_Varnakavi1

I am using gemini-2.0-flash;
In the free tier, but My api key in AI studio still shows unavailable;
and the error message is:
Tipo do erro: <class ‘google.api_core.exceptions.ResourceExhausted’>
Erro: 429 You exceeded your current quota, please check your plan and billing details. For more information on this error, head to: https://ai.google.dev/gemini-api/docs/rate-limits.

this was the only request I did today, with the prompt “Quem é o presidente do Brasil?”
only that.

I am from Brazil.

2 Likes
19

Hi bro.

  1. Model: gemini-2.5-flash
  2. Tier 1
  3. Error: [GoogleGenerativeAI Error]: Error fetching from https://generativelanguage.googleapis.com/v1beta/models/gemini-2.5-flash:generateContent: [429 Too Many Requests] You exceeded your current quota, please check your plan and billing details. For more information on this error, head to: https://ai.google.dev/gemini-api/docs/rate-limits.
    Quota exceeded for metric: generativelanguage.googleapis.com/generate_requests_per_model_per_day, limit: 0 [{“@type”:“type.googleapis.com/google.rpc.QuotaFailure",“violations”:[{“quotaMetric”:“generativelanguage.googleapis.com/generate_requests_per_model_per_day”,“quotaId”:“GenerateRequestsPerDayPerProjectPerModel”}]},{“@type”:“type.googleapis.com/google.rpc.Help”,“links”:[{“description”:"Learn more about Gemini API quotas”,“url”:“https://ai.google.dev/gemini-api/docs/rate-limits”}]}]

I have just checked about rate limit of this model and I had reached 10k RPD.

Thank for your replies.

I will upgrade to higher tier for new limitation

1 Like
21

Hi @Krish_Varnakavi1

  1. Model: All 2.5 models, including flash, flash lite, pro.

  2. Tier Free

  3. Error: The service is receiving too many requests from you

    [GoogleGenerativeAI Error]: Error fetching from https://generativelanguage.googleapis.com/v1beta/models/gemini-2.5-flash:streamGenerateContent?alt=sse: [429 Too Many Requests] You exceeded your current quota, please check your plan and billing details. For more information on this error, head to: https://ai.google.dev/gemini-api/docs/rate-limits.

22

@Krish_Varnakavi1 Please see the other thread for more details Gemini 2.5 Pro (paid, with API key) errors 100% of the time when token count is over 131k - #7 by Steven_Waterman

  1. model: “gemini-2.5-pro”
  2. Paid, happens on both API keys, both tier 3 and tier 2
  3. The error is:
{
	error: {
		code: 400,
		status: "Bad Request"
		message: {
			error: {
				code: 400,
				status: "INVALID_ARGUMENT"
				message: "The input token count (174472) exceeds the maximum number of tokens allowed (131072)."
			}
		},
	}
}

Note that the maximum tokens allowed SAYS 131072 but that is ONLY 131k tokens. Gemini 2.5 Pro support 1 MILLION tokens. So I am not over the limit.

I have DMed you with my project ID as you requested previously but you did not respond.

23

Any updates on this problem?

24

Are you guys still getting that same 429 error? Because Google Studio says everything is fine and the errors have been resolved. But for me nothing has changed. I’m still getting the 429 error.