# Feature request: Add DeepSeek V4.1 Flash 1017 to the included Cursor Models pool

**URL:** <https://forum.cursor.com/t/feature-request-add-deepseek-v4-1-flash-1017-to-the-included-cursor-models-pool/177510>\
**Category:** Feature Requests\
**Created:** [November 6, 2026, 11:08am UTC](https://forum.cursor.com/t/feature-request-add-deepseek-v4-1-flash-1017-to-the-included-cursor-models-pool/177510 "2026-11-06T11:08:24Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![pengbo](https://avatars.discourse-cdn.com/v4/letter/p/e495f1/32.png) [@pengbo](https://forum.cursor.com/u/pengbo)\
**Post date:** [November 6, 2026, 11:08am UTC](https://forum.cursor.com/t/feature-request-add-deepseek-v4-1-flash-1017-to-the-included-cursor-models-pool/177510/1 "2026-11-06T11:08:24Z")

</div>

### Feature request for product/service

AI Models

### Describe the request

Hi Cursor team and community,

I’d like to request that DeepSeek V4.1 Flash 1017 be considered for inclusion in Cursor’s internal/included model pool — i.e. the “Cursor Models” pool that uses included plan usage — rather than only being available through the paid third-party / Other Models pool.

My reasons are mainly:

Speed: In my experience, this model’s speed is excellent, especially for fast interactive coding and agent-style loops where latency matters a lot.  
Intelligence: Its intelligence is more than good enough for many everyday coding tasks, quick edits, lightweight agent workflows, and rapid iteration.  
Cost fit: As a Flash-class low-cost model, its cost profile appears potentially well-suited to Cursor’s included usage. Adding it to the internal pool could give subscribers a very attractive high-speed option without pushing them toward paid API usage.  
I understand that the included Cursor Models pool is a curated set and that Cursor bears the inference cost, so this is ultimately a product and cost decision. But I think DeepSeek V4.1 Flash 1017 could add real value to the subscription: very fast responses, good-enough intelligence for many tasks, and a likely manageable cost profile.

If full inclusion is not feasible, even a limited included quota or an experimental inclusion would be a great start.

Would the team be open to considering this? +1 from me.

Thanks!

---

<div class="post-metadata">

**Author:** ![andrewbalan](https://avatars.discourse-cdn.com/v4/letter/a/a9a28c/32.png) [@andrewbalan](https://forum.cursor.com/u/andrewbalan)\
**Post date:** [November 6, 2026, 6:10pm UTC](https://forum.cursor.com/t/feature-request-add-deepseek-v4-1-flash-1017-to-the-included-cursor-models-pool/177510/6 "2026-11-06T18:10:30Z")

</div>

+1, this would be really useful.

A fast model in the included pool is a great fit for quick edits and agent loops, where latency matters more than raw capability. It feels wasteful to use heavier models for that kind of work.

I get that it’s ultimately Cursor’s call, and there’s a cost side to it. Still, a limited quota or an experimental inclusion seems like a reasonable middle ground. It would show how people actually use the model before anyone commits to full inclusion.

If you want to check how a specific model is billed right now, it’s visible in Dashboard \> Usage. Pricing is here: [Models & Pricing | Cursor Docs](https://cursor.com/docs/models-and-pricing)
