# Gemini-1.5-pro-latest performs WORSE since yesterday. How to use its previous version? **URL:** \ **Category:** Gemini API\ **Created:** [May 15, 2024, 6:33pm UTC](https://discuss.ai.google.dev/t/gemini-1-5-pro-latest-performs-worse-since-yesterday-how-to-use-its-previous-version/2132 "2024-05-15T18:33:01Z")\ **Posts on this page:** 20\ **Page:** 1
**Author:** ![May2024](https://d1dlmcr85iqnpo.cloudfront.net/user_avatar/discuss.ai.google.dev/may2024/32/2162_2.png) [@May2024](https://discuss.ai.google.dev/u/May2024)\ **Post date:** [May 15, 2024, 6:33pm UTC](https://discuss.ai.google.dev/t/gemini-1-5-pro-latest-performs-worse-since-yesterday-how-to-use-its-previous-version/2132/1 "2024-05-15T18:33:01Z")
Hello. I’m using gemini-1.5-pro-latest for a data labeling task. The input contains images+text and the output is classification. I carefully revised my prompt, and the accuracy was around 0.78 before yesterday. However, after yesterday’s big update, when I rerun everything using the same setting, the accuracy dropped to around 0.44 suddenly. I suspect this is due to the model change yesterday. How can I use the previous version of the model “gemini-1.5-pro-latest”? GPT allows us to choose from different versions of the same model, but gemini 1.5 pro only has “latest”. I would like to go back to the previous one which gave me much higher accuracy. Looking forward to your response ASAP. Thanks! ---
**Author:** ![Macha](https://d1dlmcr85iqnpo.cloudfront.net/user_avatar/discuss.ai.google.dev/macha/32/179_2.png) [@Macha](https://discuss.ai.google.dev/u/Macha)\ **Post date:** [May 15, 2024, 9:08pm UTC](https://discuss.ai.google.dev/t/gemini-1-5-pro-latest-performs-worse-since-yesterday-how-to-use-its-previous-version/2132/2 "2024-05-15T21:08:20Z")
Hey there and welcome tot he community! What does your prompt look like? And what do the differing results look like? I’m curious to see what’s making the model borked. ---
**Author:** ![May2024](https://d1dlmcr85iqnpo.cloudfront.net/user_avatar/discuss.ai.google.dev/may2024/32/2162_2.png) [@May2024](https://discuss.ai.google.dev/u/May2024)\ **Post date:** [May 15, 2024, 9:28pm UTC](https://discuss.ai.google.dev/t/gemini-1-5-pro-latest-performs-worse-since-yesterday-how-to-use-its-previous-version/2132/3 "2024-05-15T21:28:39Z")
Thanks for your reply! The prompt is asking Gemini to be an expert in analyzing human gestures. It provides detailed descriptions and a few examples for each of 3 major gesture types. The task is to classify each gesture (based on video frames and speech text) with type and reasoning. Please see the results comparison in the attached. It performed well before (replicated multiple times to tweak the prompt) but the accuracy decreased significantly (from 0.78 to 0.43) since yesterday. Is there any way to use **the previous version of gemini-1.5-pro-latest before May 14**? Thanks! old: [[old] 0.777\_G0508\_5fps\_simple\_temp1\_V3\_replicate2 copy.xlsx - Google Sheets](https://docs.google.com/spreadsheets/d/1DalqdOC92Jn8uhSyDjQsaAOIpDGgFfSw/edit?usp=sharing&ouid=103952680812551313653&rtpof=true&sd=true) new: [[new] 0.427\_G0515\_5fps\_simple\_temp1\_V3 copy.xlsx - Google Sheets](https://docs.google.com/spreadsheets/d/1SZ6Ns6nX7E9ZFi7iSate-nq82Y2IYgr6/edit?usp=sharing&ouid=103952680812551313653&rtpof=true&sd=true) ---
**Author:** ![May2024](https://d1dlmcr85iqnpo.cloudfront.net/user_avatar/discuss.ai.google.dev/may2024/32/2162_2.png) [@May2024](https://discuss.ai.google.dev/u/May2024)\ **Post date:** [May 17, 2024, 4:13am UTC](https://discuss.ai.google.dev/t/gemini-1-5-pro-latest-performs-worse-since-yesterday-how-to-use-its-previous-version/2132/4 "2024-05-17T04:13:51Z")
Wondering if you have any suggestions? Thanks! ---
**Author:** ![Macha](https://d1dlmcr85iqnpo.cloudfront.net/user_avatar/discuss.ai.google.dev/macha/32/179_2.png) [@Macha](https://discuss.ai.google.dev/u/Macha)\ **Post date:** [May 17, 2024, 4:31am UTC](https://discuss.ai.google.dev/t/gemini-1-5-pro-latest-performs-worse-since-yesterday-how-to-use-its-previous-version/2132/5 "2024-05-17T04:31:14Z")
Yeah I’ll admit, this one is stumping me. Also it’s harder for me to read what’s in the actual cells, but I know that’s on me. The only thing I could see here being an issue is > [@May2024](#): > > (replicated multiple times to tweak the prompt) whereby the prompt tweaks might be causing issues, but still, 35% is one hell of a performance dip, even for messed up prompts. OpenAI has been infamous for stealth tweaks to their models, and so far this might be evidence of one here, but not really a solution unfortunately. And it has no issues actually seeing the video frames? You’ve proven that it’s seeing more than nothing I’m guessing? There is still a lot to learn when it comes to interacting with multimodal models because these things are so new. I wish I could provide more help 🫤. I really don’t understand why this occurred, but the data is right there proving it. Working with frontier tech is both a blessing and a curse. EDIT: Since the docs state you’ll eventually be able to specify _specific_ stable versions in the future, rest assured you’ll be able to choose the best model that fits your use case over time. One thing I noticed is that you mentioned using `gemini-1.5-pro-latest`. Go ahead and use `gemini-1.5-pro` instead. This is the stable version; the “latest” version is going to include a lot of hotfixes. That is unfortunately the best I can suggest atm. ---
**Author:** ![May2024](https://d1dlmcr85iqnpo.cloudfront.net/user_avatar/discuss.ai.google.dev/may2024/32/2162_2.png) [@May2024](https://discuss.ai.google.dev/u/May2024)\ **Post date:** [May 17, 2024, 5:37am UTC](https://discuss.ai.google.dev/t/gemini-1-5-pro-latest-performs-worse-since-yesterday-how-to-use-its-previous-version/2132/6 "2024-05-17T05:37:29Z")
Thanks so much Macha! I’ll definitely try gemini-1.5-pro to see if the stable version gives me better solution. Re: prompt, it remains exactly the same in replication. I always run (1) a “stable” prompt and (2) a “new” prompt with some tweaks. The 35% drop is based on using the same “stable” prompt and the same setting for everything. Re: seeing the video frames, I’m still working on this. Asked Gemini to output “frame description”: “briefly describe your observation for each frame, eg, img 1: … ; img 2…” but seems like some descriptions are completely off. I wonder if you have any better suggestions to best test this? Not sure if it’s because asking too many outputs may cause problems. Thank you for the additional edits! It would be great if we can specify specific stable versions in the future. I was hoping May 14 will release the stable version (they said they will start billing on May 14 for the paid API), but surprised they delayed it (again) to May 30. Wondering if you know why? ---
**Author:** ![user50](https://d1dlmcr85iqnpo.cloudfront.net/user_avatar/discuss.ai.google.dev/user50/32/1115_2.png) [@user50](https://discuss.ai.google.dev/u/user50)\ **Post date:** [May 17, 2024, 5:38am UTC](https://discuss.ai.google.dev/t/gemini-1-5-pro-latest-performs-worse-since-yesterday-how-to-use-its-previous-version/2132/7 "2024-05-17T05:38:13Z")
by the way, yes, he couldn’t even solve 1 simple example:x+x^x=1.he answered incorrectly, although the answer was 0 ---
**Author:** ![May2024](https://d1dlmcr85iqnpo.cloudfront.net/user_avatar/discuss.ai.google.dev/may2024/32/2162_2.png) [@May2024](https://discuss.ai.google.dev/u/May2024)\ **Post date:** [May 17, 2024, 5:39am UTC](https://discuss.ai.google.dev/t/gemini-1-5-pro-latest-performs-worse-since-yesterday-how-to-use-its-previous-version/2132/8 "2024-05-17T05:39:06Z")
Are you also suggesting the new 1.5 pro is worse than before? ---
**Author:** ![user50](https://d1dlmcr85iqnpo.cloudfront.net/user_avatar/discuss.ai.google.dev/user50/32/1115_2.png) [@user50](https://discuss.ai.google.dev/u/user50)\ **Post date:** [May 17, 2024, 5:41am UTC](https://discuss.ai.google.dev/t/gemini-1-5-pro-latest-performs-worse-since-yesterday-how-to-use-its-previous-version/2132/9 "2024-05-17T05:41:34Z")
I not only assume, but I am sure that he has become much dumber, but still I think that this is a bug that will be fixed within 1-2 weeks. However, if they don’t fix it, it will be very strange. ---
**Author:** ![May2024](https://d1dlmcr85iqnpo.cloudfront.net/user_avatar/discuss.ai.google.dev/may2024/32/2162_2.png) [@May2024](https://discuss.ai.google.dev/u/May2024)\ **Post date:** [May 17, 2024, 5:53am UTC](https://discuss.ai.google.dev/t/gemini-1-5-pro-latest-performs-worse-since-yesterday-how-to-use-its-previous-version/2132/10 "2024-05-17T05:53:20Z")
Thanks for sharing your experience! Feel better it was not just me… Let’s hope they will fix the bug within 1-2 weeks! Hope some folks working on this at Google will see our posts!!! ---
**Author:** ![OrangiaNebula](https://d1dlmcr85iqnpo.cloudfront.net/user_avatar/discuss.ai.google.dev/orangianebula/32/480_2.png) [@OrangiaNebula](https://discuss.ai.google.dev/u/OrangiaNebula)\ **Post date:** [May 17, 2024, 6:02am UTC](https://discuss.ai.google.dev/t/gemini-1-5-pro-latest-performs-worse-since-yesterday-how-to-use-its-previous-version/2132/11 "2024-05-17T06:02:09Z")
On the mathematical capabilities, I can confirm that the newly released 1.5 model degraded compared to the predecessor: > [@A thread on Mathematics](http://discuss.ai.google.dev/t/a-thread-on-mathematics/1464/13): > > Clear and comprehensive. And, best I can tell, GPT-4o didn’t even have to use Wolfram to generate the proof. And as to whether our concerns are getting attention, I am worried that a flood of questions about the developer competition is depriving all the oxygen, and nobody from Google is listening to us. ---
**Author:** ![May2024](https://d1dlmcr85iqnpo.cloudfront.net/user_avatar/discuss.ai.google.dev/may2024/32/2162_2.png) [@May2024](https://discuss.ai.google.dev/u/May2024)\ **Post date:** [May 17, 2024, 6:06am UTC](https://discuss.ai.google.dev/t/gemini-1-5-pro-latest-performs-worse-since-yesterday-how-to-use-its-previous-version/2132/12 "2024-05-17T06:06:23Z")
Thank you for sharing this! It is really disappointing. I switched from GPT to Gemini because of its better visual understanding. But this recent upgrade is really a bummer. I’ll keep posting and hopefully someone will pay attention. Maybe they will finally catch up when they start billing on May 30… ---
**Author:** ![May2024](https://d1dlmcr85iqnpo.cloudfront.net/user_avatar/discuss.ai.google.dev/may2024/32/2162_2.png) [@May2024](https://discuss.ai.google.dev/u/May2024)\ **Post date:** [May 17, 2024, 6:14am UTC](https://discuss.ai.google.dev/t/gemini-1-5-pro-latest-performs-worse-since-yesterday-how-to-use-its-previous-version/2132/13 "2024-05-17T06:14:34Z")
I just tried it and seems like it only has `gemini-1.5-pro-latest` instead of `gemini-1.5-pro`. Maybe I missed something, is there any way to use `gemini-1.5-pro`? ---
**Author:** ![OrangiaNebula](https://d1dlmcr85iqnpo.cloudfront.net/user_avatar/discuss.ai.google.dev/orangianebula/32/480_2.png) [@OrangiaNebula](https://discuss.ai.google.dev/u/OrangiaNebula)\ **Post date:** [May 17, 2024, 6:31am UTC](https://discuss.ai.google.dev/t/gemini-1-5-pro-latest-performs-worse-since-yesterday-how-to-use-its-previous-version/2132/14 "2024-05-17T06:31:29Z")
We can only use what list\_models() tells us is there. I get these: ```auto models/chat-bison-001 1024 models/text-bison-001 1024 models/embedding-gecko-001 1 models/gemini-1.0-pro 2048 models/gemini-1.0-pro-001 2048 models/gemini-1.0-pro-latest 2048 models/gemini-1.0-pro-vision-latest 4096 models/gemini-1.5-flash-latest 8192 models/gemini-1.5-pro-latest 8192 models/gemini-pro 2048 models/gemini-pro-vision 4096 models/embedding-001 1 models/text-embedding-004 1 models/aqa 1024 ``` The second column is maximum tokens returned, don’t let it confuse you ---
**Author:** ![May2024](https://d1dlmcr85iqnpo.cloudfront.net/user_avatar/discuss.ai.google.dev/may2024/32/2162_2.png) [@May2024](https://discuss.ai.google.dev/u/May2024)\ **Post date:** [May 17, 2024, 6:45am UTC](https://discuss.ai.google.dev/t/gemini-1-5-pro-latest-performs-worse-since-yesterday-how-to-use-its-previous-version/2132/15 "2024-05-17T06:45:48Z")
Thank you! Seems like gemini-1.5-pro provides only the “latest” version… Google should at least provide both “old” and “new” versions if it’s really a big update (e.g., gpt4 has so many different versions to choose, with dates in names) ---
**Author:** ![Macha](https://d1dlmcr85iqnpo.cloudfront.net/user_avatar/discuss.ai.google.dev/macha/32/179_2.png) [@Macha](https://discuss.ai.google.dev/u/Macha)\ **Post date:** [May 17, 2024, 7:00pm UTC](https://discuss.ai.google.dev/t/gemini-1-5-pro-latest-performs-worse-since-yesterday-how-to-use-its-previous-version/2132/16 "2024-05-17T19:00:17Z")
Ah, my bad. I was looking at these docs when trying to answer your question: > **[Gemini models  |  Gemini API  |  Google AI for Developers](https://ai.google.dev/gemini-api/docs/models/gemini#gemini-1.5-pro)** > > Learn about Google's most advanced AI models, the Gemini model family, including Gemini 1.5 Flash, Gemini 1.5 Pro, and more In _there_, they state the latest stable version is `gemini-1.5-pro`, but as @OrangiaNebula points out, that’s I guess not on the list using list\_models(). So, looks like we found a discrepancy between what’s in the docs and what’s currently out there. ---
**Author:** ![OrangiaNebula](https://d1dlmcr85iqnpo.cloudfront.net/user_avatar/discuss.ai.google.dev/orangianebula/32/480_2.png) [@OrangiaNebula](https://discuss.ai.google.dev/u/OrangiaNebula)\ **Post date:** [May 17, 2024, 10:31pm UTC](https://discuss.ai.google.dev/t/gemini-1-5-pro-latest-performs-worse-since-yesterday-how-to-use-its-previous-version/2132/17 "2024-05-17T22:31:23Z")
> [@May2024](#): > > Seems like gemini-1.5-pro provides only the “latest” version… Google should at least provide both “old” and “new” versions if it’s really a big update Hi, @Lloyd_Hightower , I believe you are an operations person and will likely appreciate the wisdom of the proposal of participant May 2024. ---
**Author:** ![afirstenberg](https://d1dlmcr85iqnpo.cloudfront.net/user_avatar/discuss.ai.google.dev/afirstenberg/32/202_2.png) [@afirstenberg](https://discuss.ai.google.dev/u/afirstenberg)\ **Post date:** [May 18, 2024, 4:27am UTC](https://discuss.ai.google.dev/t/gemini-1-5-pro-latest-performs-worse-since-yesterday-how-to-use-its-previous-version/2132/18 "2024-05-18T04:27:39Z")
> [@Macha](#): > > In _there_, they state the latest stable version is `gemini-1.5-pro`, Yeah, but they _also_ state that Gemini 1.5 is a preview version. And if you look down at [Gemini API  |  Google for Developers](https://ai.google.dev/gemini-api/docs/models/gemini#model-versions) it says > Gemini models are available in either _preview_ or _stable_ versions. This suggests, in typical Google confusing documentation, that there **is no stable version** of Gemini 1.5. So only the model with the `-latest` suffix is valid. Even tho the documentation says what the stable version shortcut name would be. Note that they don’t list a version with a version number suffix, which would indicate a stable version. This is also why there is an unexpected change. They don’t say it’s stable. ---
**Author:** ![May2024](https://d1dlmcr85iqnpo.cloudfront.net/user_avatar/discuss.ai.google.dev/may2024/32/2162_2.png) [@May2024](https://discuss.ai.google.dev/u/May2024)\ **Post date:** [May 18, 2024, 7:06am UTC](https://discuss.ai.google.dev/t/gemini-1-5-pro-latest-performs-worse-since-yesterday-how-to-use-its-previous-version/2132/19 "2024-05-18T07:06:21Z")
Thank you for helping connect with the right person! Really hope they could fix the problem soon 🙏 I preferred Gemini over GPT in the last couple of weeks due to its better performance in visual understanding/reasoning. This recent major update on May 14 is really disapointing 😭 ---
**Author:** ![May2024](https://d1dlmcr85iqnpo.cloudfront.net/user_avatar/discuss.ai.google.dev/may2024/32/2162_2.png) [@May2024](https://discuss.ai.google.dev/u/May2024)\ **Post date:** [May 18, 2024, 7:17am UTC](https://discuss.ai.google.dev/t/gemini-1-5-pro-latest-performs-worse-since-yesterday-how-to-use-its-previous-version/2132/20 "2024-05-18T07:17:33Z")
@Logan_Kilpatrick Hi Logan! Hope all is well! Just saw your reply in another post which is super helpful and thought you might be the right person to ask. Wondering if you could help me with this question here about Gemini when you get a chance. Thanks in advance for your time! [Next page](https://discuss.ai.google.dev/t/gemini-1-5-pro-latest-performs-worse-since-yesterday-how-to-use-its-previous-version/2132.md?page=2)