Am I not considered an active user? Why did my 2.5 Pro option disappear? What are your black-box criteria for this?
I subscribed specifically to use 2.5 Pro. Removing this model without any warning borders on bait-and-switch and consumer fraud.
I am an active user of 2.5 Pro. Bring my 2.5 Pro option back to AI Studio!
I completely agree that different models have their own strengths. I hope you will not deprecate Gemini 2.5 Pro, or at least open-source it!
I agree with you, paid subscribers should at least get 2.5 pro back. 3.1 pro was a disaster.
Hello @Kiri @Total_Anarchy ,
To ensure reliable performance for everyone, we are limiting access to the 2.5 models to users who have actively used them in the past. These models are not deprecated and will continue to be served until further notice.
This restriction now also applies to Google AI Studio, where the gemini-2.5-xxx are no longer available.
For your new projects, please use our latest models: 3.5 Flash-Lite or 3.8 Flash. This helps us maintain sufficient capacity for both ongoing legacy workflows and new applications.
Your response directly contradicts itself.
If you claim that:
* Access is being “limited to users who have actively used them in the past”, and
* “This restriction now also applies to Google AI Studio”
Then by your own stated policy, active legacy users should still retain access in AI Studio.
I have been making high-volume, daily calls to 2.5 Pro. By any reasonable technical standard, I am an active user. Why has the 2.5 Pro option completely disappeared from my AI Studio UI? What is the exact, transparent metric you use to define an “active user,” or is this an account whitelisting bug that needs to be resolved for active accounts?
Furthermore, suggesting 3.5 Flash-Lite or 3.8 Flash as alternatives to 2.5 Pro is completely disconnected from reality. Across numerous practical use cases, these newer models simply do not measure up to 2.5 Pro—they perform noticeably worse, and they are nowhere near an adequate replacement. Our production pipelines, prompt architectures, and workflows were engineered specifically around 2.5 Pro’s outputs.
You open your message claiming this is “to ensure reliable performance”, yet you just crippled the stability and availability of your most active users without warning. Offloading platform capacity shortages onto established users is unacceptable.
Please provide a genuine, technical answer rather than recycling this canned macro: How do active, qualifying users get their 2.5 Pro UI access restored in Google AI Studio?
What exactly qualifies as an “actively used” user?
I was actively using Gemini 2.5 Pro in AI Studio, including just four days before it disappeared, yet I still lost access.
So what counts as an active user? What is the actual threshold?
Or is this simply an excuse for removing Gemini 2.5 Pro without any prior notice?
Right now, it feels like there are no “active users” at all, and this explanation is just being used to justify an unexpected removal after the fact.
If there really are users who still have access, please clearly explain the criteria instead of making vague statements.
Removing Gemini 2.5 Pro without any warning is completely unacceptable. You didn’t just remove access to Gemini 2.5 Pro in AI Studio—you also took away the ability to continue using existing 2.5 Pro chats and even removed access through the API.
The official retirement date still says “TBD.” If that’s the case, why was it removed in such a sudden and disruptive way that affects everyone?
Even if you needed to increase the price, or require users to pay for API usage themselves, that would have been far more acceptable than taking it away without warning. Did anyone consider that, for some of us, our Gemini 2.5 Pro conversations and the model itself are deeply important and simply cannot be replaced?
Most people here are paying Gemini users. We ask Google to restore Gemini 2.5 Pro and, at the very least, provide a minimum of 2–4 weeks’ advance notice before removing a model, so users have time to prepare and say goodbye to their ongoing conversations.
To be honest, Gemini 2.5 Pro is the main reason many of us choose Google’s AI products over the alternatives.
Exactly! 2.5 Pro is the very reason so many of us use Google AI products!
I am an active user of this model. What criteria are you actually using for “active users?” Because Gemini 3.1 Pro is effectively useless right now. Literally no one can access it reliably, and it’s breaking production pipelines across the board.
I know Gemini 4 Pro is in post-training right now, and that process is compute-intensive. RLHF, RLAIF, and safety evaluation phases require the same TPU pod architecture as inference. If Google can’t run Gemini 2.5 Pro and 3.x model inference at full scale, and 4 Pro post-training on fixed hardware, then say it formally. Don’t sidestep paying users with vague corporate language and restrictions.
The team already knows that deprecating 2.5 Pro caused enough community backlash that you reversed the October date. This “active user” restriction achieves roughly the same TPU reallocation without triggering a formal deprecation announcement. You need to be honest about what’s actually happening here. The user base is filling up with angry developers, and the lack of transparency is making it worse.
Either acknowledge the hardware constraint or explain the real criteria. Corporate non-answers won’t cut it when people’s production systems are down without a good replacement.
Yes Brother. They should have at least given a advanced notice regarding this change! It shows What they have thought of their consumers in a AI Game, According to them we are just Pawns.
This type of authoritarianism at google will cost it’s legacy! The company thinks it is too big to fail, but history shows the opposite!
There is no official announcement from Google about Gemini 2.5 pro deprecation date.
We pay for the access to 2.5 pro model and it’s absolutely inappropriate behaviour on Google side to switch it off silently.
Please, open more tickets, address @Logan_Kilpatrick via email or on X.
We should draw as much attention as possible to this problem.
Yes. Exactly. This is such an easily avoided, self-inflicted wound. I wish I could say I was surprised. Plenty of us lined up here with wallets open, but if we’re not enterprise, we don’t warrant any notice I suppose.