# What model sits between Qwen 3.8 27b and Flash next for coding? — retained record

> Retained reddit record from reddit:LocalLLaMA.

Canonical: https://highsignal.app/data/records/f0001daaf0de138d

[← sources](https://highsignal.app/data)


source: [reddit](https://highsignal.app/data/reddit) (reddit:LocalLLaMA)  published 2026-09-28 13:25 ingested 2026-09-29 02:43

[original source ↗](https://www.reddit.com/r/LocalLLaMA/comments/1wsekkj/what_model_sits_between_qwen_38_27b_and_flash/)

## retained text



```
Having tested both Qwen 3.8 27b and Flash next on RTX 5090 with 96GB RAM, I want to find the middle ground between the two for coding capabilities but not sacrifice decode speed to standstill. I would like decode speed to between 75-100 ideally for fast iterations; otherwise I become impatient. 

Currently I get 200+ TPS on Qwen 3.8 27B and approx 50 TPS on Flash next.

My hardware - RTX 5090 and 96 GB DDR5 which I plan to upgrade to 128 GB (in this economy, yes, but unwillingly).

What model sits between these two in terms of coding capabilities and hardware requirement? If none is present, I can perhaps think of using Flash next for plan creation and 27b for implementation.
```



document fetched 2026-09-29 00:43 · canonical https://reddit.com/r/LocalLLaMA/comments/1wsekkj/what_model_sits_between_qwen_38_27b_and_flash
