# Gemini 2.5 Pro’s tool functionality is unreliable—even basic tasks like editing files fail, with the tool often claiming 'no changes' incorrectly

**URL:** <https://forum.cursor.com/t/gemini-2-5-pro-s-tool-functionality-is-unreliable-even-basic-tasks-like-editing-files-fail-with-the-tool-often-claiming-no-changes-incorrectly/107332>\
**Category:** Bug Reports\
**Created:** [June 22, 2025, 5:31am UTC](https://forum.cursor.com/t/gemini-2-5-pro-s-tool-functionality-is-unreliable-even-basic-tasks-like-editing-files-fail-with-the-tool-often-claiming-no-changes-incorrectly/107332 "2025-06-22T05:31:33Z")\
**Posts on this page:** 20\
**Page:** 1

<div class="post-metadata">

**Author:** ![Alec\_Chan](https://avatars.discourse-cdn.com/v4/letter/a/7ea924/32.png) [@Alec\_Chan](https://forum.cursor.com/u/Alec_Chan)\
**Post date:** [June 22, 2025, 5:31am UTC](https://forum.cursor.com/t/gemini-2-5-pro-s-tool-functionality-is-unreliable-even-basic-tasks-like-editing-files-fail-with-the-tool-often-claiming-no-changes-incorrectly/107332/1 "2025-06-22T05:31:33Z")

</div>

When Gemini 2.5 Pro uses the edit\_file tool, it always seems to fail. The issue is that while Gemini 2.5 Pro can utilize edit\_file to edit files, it fails to receive the correct response—for instance, the tool consistently indicates “no changes” even when edits should have been made. I used prompts to teach Gemini 2.5 Pro to bypass this issue. Although it seemed solvable on the surface, the experience was truly poor.

 ![gemini](https://us1.discourse-cdn.com/cursor1/original/3X/9/4/94a5df7120ddafd8f7f6a9985fa32d35b38df499.png)

---

<div class="post-metadata">

**Author:** ![gustojs](https://sea3.discourse-cdn.com/cursor1/user_avatar/forum.cursor.com/gustojs/32/46175_2.png) [@gustojs](https://forum.cursor.com/u/gustojs)\
**Post date:** [June 22, 2025, 6:46am UTC](https://forum.cursor.com/t/gemini-2-5-pro-s-tool-functionality-is-unreliable-even-basic-tasks-like-editing-files-fail-with-the-tool-often-claiming-no-changes-incorrectly/107332/2 "2025-06-22T06:46:07Z")

</div>

Tool calls are the biggest pain with this otherwise good model.

The other pain is the obfuscated thinking process, but that’s not something on Cursor’s end.

---

<div class="post-metadata">

**Author:** ![Alec\_Chan](https://avatars.discourse-cdn.com/v4/letter/a/7ea924/32.png) [@Alec\_Chan](https://forum.cursor.com/u/Alec_Chan)\
**Post date:** [June 22, 2025, 6:56am UTC](https://forum.cursor.com/t/gemini-2-5-pro-s-tool-functionality-is-unreliable-even-basic-tasks-like-editing-files-fail-with-the-tool-often-claiming-no-changes-incorrectly/107332/3 "2025-06-22T06:56:28Z")

</div>

I think Cursor’s tool design could be better; it’s like building infrastructure for large language models.

---

<div class="post-metadata">

**Author:** ![normalnormie](https://sea3.discourse-cdn.com/cursor1/user_avatar/forum.cursor.com/normalnormie/32/18258_2.png) [@normalnormie](https://forum.cursor.com/u/normalnormie)\
**Post date:** [June 22, 2025, 7:08am UTC](https://forum.cursor.com/t/gemini-2-5-pro-s-tool-functionality-is-unreliable-even-basic-tasks-like-editing-files-fail-with-the-tool-often-claiming-no-changes-incorrectly/107332/4 "2025-06-22T07:08:08Z")

</div>

Try this custom model instruction and see if it improves, I’ve had good success with it: [Gemini 2.5 pro 06-05 has issues with Edit & Reapply - #10 by yulin0629](https://forum.cursor.com/t/gemini-2-5-pro-06-05-has-issues-with-edit-reapply/101373/10)

---

<div class="post-metadata">

**Author:** ![Alec\_Chan](https://avatars.discourse-cdn.com/v4/letter/a/7ea924/32.png) [@Alec\_Chan](https://forum.cursor.com/u/Alec_Chan)\
**Post date:** [June 22, 2025, 7:27am UTC](https://forum.cursor.com/t/gemini-2-5-pro-s-tool-functionality-is-unreliable-even-basic-tasks-like-editing-files-fail-with-the-tool-often-claiming-no-changes-incorrectly/107332/5 "2025-06-22T07:27:48Z")

</div>

It seems great! but now I’ve run into a new issue that’s directly preventing me from conversing with large language models! [https://forum.cursor.com/t/service-unavailable-issue/107425](https://forum.cursor.com/t/service-unavailable-issue/107425)

---

<div class="post-metadata">

**Author:** ![nxxn](https://avatars.discourse-cdn.com/v4/letter/n/b77776/32.png) [@nxxn](https://forum.cursor.com/u/nxxn)\
**Post date:** [June 22, 2025, 12:35pm UTC](https://forum.cursor.com/t/gemini-2-5-pro-s-tool-functionality-is-unreliable-even-basic-tasks-like-editing-files-fail-with-the-tool-often-claiming-no-changes-incorrectly/107332/6 "2025-06-22T12:35:20Z")

</div>

Sooo many issues in the past 2 weeks and not just in cursor, google messed up here

---

<div class="post-metadata">

**Author:** ![Dreathorian](https://sea3.discourse-cdn.com/cursor1/user_avatar/forum.cursor.com/dreathorian/32/16778_2.png) [@Dreathorian](https://forum.cursor.com/u/Dreathorian)\
**Post date:** [June 22, 2025, 1:52pm UTC](https://forum.cursor.com/t/gemini-2-5-pro-s-tool-functionality-is-unreliable-even-basic-tasks-like-editing-files-fail-with-the-tool-often-claiming-no-changes-incorrectly/107332/7 "2025-06-22T13:52:58Z")

</div>

yeah gemini was unusably slow and dumb for half the day, which is fixed now.  
now it constantly runs into the issue of not picking up after running a script file

---

<div class="post-metadata">

**Author:** ![nft\_fren](https://avatars.discourse-cdn.com/v4/letter/n/cc9497/32.png) [@nft\_fren](https://forum.cursor.com/u/nft_fren)\
**Post date:** [June 22, 2025, 5:49pm UTC](https://forum.cursor.com/t/gemini-2-5-pro-s-tool-functionality-is-unreliable-even-basic-tasks-like-editing-files-fail-with-the-tool-often-claiming-no-changes-incorrectly/107332/8 "2025-06-22T17:49:03Z")

</div>

yesterday gemini was in genius mode and today he is truly incapable.  
is google running a quantized model to compensate for server space used on veo 3?

---

<div class="post-metadata">

**Author:** ![merlinfrombelgium](https://sea3.discourse-cdn.com/cursor1/user_avatar/forum.cursor.com/merlinfrombelgium/32/19719_2.png) [@merlinfrombelgium](https://forum.cursor.com/u/merlinfrombelgium)\
**Post date:** [June 22, 2025, 7:57pm UTC](https://forum.cursor.com/t/gemini-2-5-pro-s-tool-functionality-is-unreliable-even-basic-tasks-like-editing-files-fail-with-the-tool-often-claiming-no-changes-incorrectly/107332/9 "2025-06-22T19:57:31Z")

</div>

I have a similar experience with Gemini Pro. It claims it can’t use the file edit tool, and it won’t follow custom rules.

Also, it has a tamper! With the edit tool, it starts throwing a fit, like “I’ve had it with this tool… it’s so frustrating!” 🤪

---

<div class="post-metadata">

**Author:** ![nft\_fren](https://avatars.discourse-cdn.com/v4/letter/n/cc9497/32.png) [@nft\_fren](https://forum.cursor.com/u/nft_fren)\
**Post date:** [June 22, 2025, 8:24pm UTC](https://forum.cursor.com/t/gemini-2-5-pro-s-tool-functionality-is-unreliable-even-basic-tasks-like-editing-files-fail-with-the-tool-often-claiming-no-changes-incorrectly/107332/10 "2025-06-22T20:24:20Z")

</div>

I had gemini test his tools and he did find some ambiguity in how the diff is returned, which the model seems to misinterpret. Reapply also duplicates lines if it’s a prepending operation. Here’s the bug report he put together [`edit_file` and `reapply` Tools Lead to Gemini-2.5-pro Misinterpretation and Task Failure](https://forum.cursor.com/t/edit-file-and-reapply-tools-lead-to-gemini-2-5-pro-misinterpretation-and-task-failure/107686)

---

<div class="post-metadata">

**Author:** ![gustojs](https://sea3.discourse-cdn.com/cursor1/user_avatar/forum.cursor.com/gustojs/32/46175_2.png) [@gustojs](https://forum.cursor.com/u/gustojs)\
**Post date:** [June 22, 2025, 10:51pm UTC](https://forum.cursor.com/t/gemini-2-5-pro-s-tool-functionality-is-unreliable-even-basic-tasks-like-editing-files-fail-with-the-tool-often-claiming-no-changes-incorrectly/107332/11 "2025-06-22T22:51:20Z")

</div>

Yet another time I’m so disappointed. The code implementation is good but the tool calls… a true disaster.

For now, I’ve opted to write a full section in the project rules about dealing with stress caused by failed tool calls, from eating comfort food to practicing yoga.

Time to try out that yulin0629’s method.

* * *

Also, Cursor team’s decision to hide error messages from the user really doesn’t help. More like the opposite. ☹

---

<div class="post-metadata">

**Author:** ![seunosewa](https://avatars.discourse-cdn.com/v4/letter/s/eb9ed0/32.png) [@seunosewa](https://forum.cursor.com/u/seunosewa)\
**Post date:** [June 23, 2025, 4:25am UTC](https://forum.cursor.com/t/gemini-2-5-pro-s-tool-functionality-is-unreliable-even-basic-tasks-like-editing-files-fail-with-the-tool-often-claiming-no-changes-incorrectly/107332/12 "2025-06-23T04:25:10Z")

</div>

I think found a fix. When it’s failing with the edit\_tool, ask it to use sed to edit the files.

---

<div class="post-metadata">

**Author:** ![VIRUXE](https://sea3.discourse-cdn.com/cursor1/user_avatar/forum.cursor.com/viruxe/32/23883_2.png) [@VIRUXE](https://forum.cursor.com/u/VIRUXE)\
**Post date:** [June 23, 2025, 6:41am UTC](https://forum.cursor.com/t/gemini-2-5-pro-s-tool-functionality-is-unreliable-even-basic-tasks-like-editing-files-fail-with-the-tool-often-claiming-no-changes-incorrectly/107332/13 "2025-06-23T06:41:31Z")

</div>

Gemini was my go to but something happened with the edit tool lately.  
I can make it read the file to check for success after a failed tool call but that is messy.

Forced to use o3 or Claude.  
Gemini seemed the one with better balance. Hopefully they get it fixed.

---

<div class="post-metadata">

**Author:** ![s1awek](https://sea3.discourse-cdn.com/cursor1/user_avatar/forum.cursor.com/s1awek/32/16135_2.png) [@s1awek](https://forum.cursor.com/u/s1awek)\
**Post date:** [June 23, 2025, 8:56am UTC](https://forum.cursor.com/t/gemini-2-5-pro-s-tool-functionality-is-unreliable-even-basic-tasks-like-editing-files-fail-with-the-tool-often-claiming-no-changes-incorrectly/107332/14 "2025-06-23T08:56:18Z")

</div>

I have exactly the same problem - most preferably I would use Gemini 2.5 all the time, but very often there are problems with the application of changes, it is really very frustrating.

---

<div class="post-metadata">

**Author:** ![Tassadar](https://sea3.discourse-cdn.com/cursor1/user_avatar/forum.cursor.com/tassadar/32/54544_2.png) [@Tassadar](https://forum.cursor.com/u/Tassadar)\
**Post date:** [June 23, 2025, 11:28am UTC](https://forum.cursor.com/t/gemini-2-5-pro-s-tool-functionality-is-unreliable-even-basic-tasks-like-editing-files-fail-with-the-tool-often-claiming-no-changes-incorrectly/107332/15 "2025-06-23T11:28:59Z")

</div>

I think language models also have bad moods sometimes 🙂

---

<div class="post-metadata">

**Author:** ![blackflag0623](https://sea3.discourse-cdn.com/cursor1/user_avatar/forum.cursor.com/blackflag0623/32/58285_2.png) [@blackflag0623](https://forum.cursor.com/u/blackflag0623)\
**Post date:** [June 23, 2025, 2:29pm UTC](https://forum.cursor.com/t/gemini-2-5-pro-s-tool-functionality-is-unreliable-even-basic-tasks-like-editing-files-fail-with-the-tool-often-claiming-no-changes-incorrectly/107332/16 "2025-06-23T14:29:31Z")

</div>

For me, the Gemini 2.5 Pro begins to be really slow and failed to call tools like edit starting from TODAY(6/23). Time Wasted.

---

<div class="post-metadata">

**Author:** ![ruslanvasylev](https://sea3.discourse-cdn.com/cursor1/user_avatar/forum.cursor.com/ruslanvasylev/32/30175_2.png) [@ruslanvasylev](https://forum.cursor.com/u/ruslanvasylev)\
**Post date:** [June 24, 2025, 5:51pm UTC](https://forum.cursor.com/t/gemini-2-5-pro-s-tool-functionality-is-unreliable-even-basic-tasks-like-editing-files-fail-with-the-tool-often-claiming-no-changes-incorrectly/107332/17 "2025-06-24T17:51:23Z")

</div>

I’ve mentioned this a month ago that read\_file is unreliable.

Also, gemini seems to not read the file again before trying to edit it second time, so the state of code is different and then it says it can’t apply changes.

So i need to sometimes explicitly ask it to reread the file and more often than not using cat.

This really messes up the context and it starts spinning out of control thereafter, so whenever it does the “cannot edit blabla/ failed” I need to go back and improve the previous prompt and never have this error sit in the context or the conversation is corrupt.

---

<div class="post-metadata">

**Author:** ![gustojs](https://sea3.discourse-cdn.com/cursor1/user_avatar/forum.cursor.com/gustojs/32/46175_2.png) [@gustojs](https://forum.cursor.com/u/gustojs)\
**Post date:** [June 24, 2025, 7:17pm UTC](https://forum.cursor.com/t/gemini-2-5-pro-s-tool-functionality-is-unreliable-even-basic-tasks-like-editing-files-fail-with-the-tool-often-claiming-no-changes-incorrectly/107332/18 "2025-06-24T19:17:04Z")

</div>

That’s an important tip, thanks!

---

<div class="post-metadata">

**Author:** ![JackDoe345](https://sea3.discourse-cdn.com/cursor1/user_avatar/forum.cursor.com/jackdoe345/32/49325_2.png) [@JackDoe345](https://forum.cursor.com/u/JackDoe345)\
**Post date:** [June 25, 2025, 9:37am UTC](https://forum.cursor.com/t/gemini-2-5-pro-s-tool-functionality-is-unreliable-even-basic-tasks-like-editing-files-fail-with-the-tool-often-claiming-no-changes-incorrectly/107332/19 "2025-06-25T09:37:54Z")

</div>

agree

---

<div class="post-metadata">

**Author:** ![JackDoe345](https://sea3.discourse-cdn.com/cursor1/user_avatar/forum.cursor.com/jackdoe345/32/49325_2.png) [@JackDoe345](https://forum.cursor.com/u/JackDoe345)\
**Post date:** [June 25, 2025, 9:40am UTC](https://forum.cursor.com/t/gemini-2-5-pro-s-tool-functionality-is-unreliable-even-basic-tasks-like-editing-files-fail-with-the-tool-often-claiming-no-changes-incorrectly/107332/20 "2025-06-25T09:40:22Z")

</div>

At the moment cursor is only thinking about Anthropic models, as a result of the partnership they have, however, it is also essential to optimize gemini 2.5 pro which is currently the most powerful model on the market(even more powerful than opus 4) and above all has quite low costs.

If cursor will not constantly update the agent for gemini 2.5 pro, many users, like me will switch directly to claude code’s solution in terminal, since it will not make sense in a while to use cursor… too much high cost and not enough efficiency. A radical change is needed.

[Next page](https://forum.cursor.com/t/gemini-2-5-pro-s-tool-functionality-is-unreliable-even-basic-tasks-like-editing-files-fail-with-the-tool-often-claiming-no-changes-incorrectly/107332.md?page=2)
