# New video: DeepSeek V4.1 Flash on Strix Halo

**URL:** <https://community.frame.work/t/new-video-deepseek-v4-1-flash-on-strix-halo/85033>\
**Category:** Framework Desktop\
**Created:** [September 23, 2026, 11:45am UTC](https://community.frame.work/t/new-video-deepseek-v4-1-flash-on-strix-halo/85033 "2026-09-23T11:45:49Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![kyuz0](https://avatars.discourse-cdn.com/v4/letter/k/b77776/32.png) [@kyuz0](https://community.frame.work/u/kyuz0)\
**Post date:** [September 23, 2026, 11:45am UTC](https://community.frame.work/t/new-video-deepseek-v4-1-flash-on-strix-halo/85033/1 "2026-09-23T11:45:50Z")

</div>

I spent the last week adding support to DwarfStar and testing two ways to run it: SSD streaming on one 128 GB machine, or distributing the model across two nodes over TCP or RoCE.

[![](https://us1.discourse-cdn.com/flex001/uploads/framework3/original/3X/4/3/43a301694f7441eed7f0d4ea9313625e597c07e6.jpeg "DeepSeek V4.1 Flash on Strix Halo: Single-Node SSD vs Two-Node Cluster (DwarfStar)") ](https://www.youtube.com/watch?v=1DaMkTuiCEQ)

---

<div class="post-metadata">

**Author:** ![Kai\_Fricke](https://avatars.discourse-cdn.com/v4/letter/k/5f8ce5/32.png) [@Kai\_Fricke](https://community.frame.work/u/Kai_Fricke)\
**Post date:** [September 23, 2026, 5:01pm UTC](https://community.frame.work/t/new-video-deepseek-v4-1-flash-on-strix-halo/85033/2 "2026-09-23T17:01:31Z")

</div>

Great video as usual! Did I miss a comparison to the Halogen-Flash server?

I only bought one machine and now they are far above my budget.  
So how are the tradeoffs to the smaller Qwen3.8-Flash-Next model?

---

<div class="post-metadata">

**Author:** ![kyuz0](https://avatars.discourse-cdn.com/v4/letter/k/b77776/32.png) [@kyuz0](https://community.frame.work/u/kyuz0)\
**Post date:** [September 23, 2026, 5:38pm UTC](https://community.frame.work/t/new-video-deepseek-v4-1-flash-on-strix-halo/85033/3 "2026-09-23T17:38:01Z")

</div>

I made a separate video on Halogen, ultimately it’s a different model also in size, Qwen 3.8 Flash is faster, this will be more capable overall, but you’ll get a lot done with Qwen 3.8 Flash Next, it’s probably one of the best models for a single Strix Halo right now.