# Add Gemini 3 Pro High as a model in Cursor

**URL:** <https://forum.cursor.com/t/add-gemini-3-pro-high-as-a-model-in-cursor/143302>\
**Category:** Feature Requests\
**Created:** [November 19, 2025, 8:08pm UTC](https://forum.cursor.com/t/add-gemini-3-pro-high-as-a-model-in-cursor/143302 "2025-11-19T20:08:14Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![Kazopl](https://sea3.discourse-cdn.com/cursor1/user_avatar/forum.cursor.com/kazopl/32/90359_2.png) [@Kazopl](https://forum.cursor.com/u/Kazopl)\
**Post date:** [November 19, 2025, 8:08pm UTC](https://forum.cursor.com/t/add-gemini-3-pro-high-as-a-model-in-cursor/143302/1 "2025-11-19T20:08:14Z")

</div>

### Feature request for product/service

AI Models

### Describe the request

It would be amazing to have full support for Gemini 3 Pro High in Cursor as soon as possible.

### Operating System (if it applies)

Windows 10/11  
MacOS  
Linux

---

<div class="post-metadata">

**Author:** ![liquefy](https://avatars.discourse-cdn.com/v4/letter/l/edb3f5/32.png) [@liquefy](https://forum.cursor.com/u/liquefy)\
**Post date:** [November 19, 2025, 8:13pm UTC](https://forum.cursor.com/t/add-gemini-3-pro-high-as-a-model-in-cursor/143302/2 "2025-11-19T20:13:36Z")

</div>

i think it works as high

 ![image](https://us1.discourse-cdn.com/cursor1/original/3X/a/2/a2d84280183cb83753b72c2f208528bd7f12e29b.png)

---

<div class="post-metadata">

**Author:** ![Kazopl](https://sea3.discourse-cdn.com/cursor1/user_avatar/forum.cursor.com/kazopl/32/90359_2.png) [@Kazopl](https://forum.cursor.com/u/Kazopl)\
**Post date:** [November 19, 2025, 8:31pm UTC](https://forum.cursor.com/t/add-gemini-3-pro-high-as-a-model-in-cursor/143302/3 "2025-11-19T20:31:24Z")

</div>

> [@liquefy](#):
>
> i think it works as high

 ![image](https://us1.discourse-cdn.com/cursor1/original/3X/7/d/7db79cd1f9e466e8275659be164cc4b29be91df1.png)

By default, the model uses dynamic reasoning, which means it adjusts its depth of thinking on the fly. If you want consistently deeper reasoning instead of whatever the model decides, you have to explicitly set the thinking level to High. [Link here](https://ai.google.dev/gemini-api/docs/gemini-3?thinking=high)