# Quick One Shot Pelican RTX 5090 Test - Qwen 3.8 27B. Q4_K_M vs Q5_K_M Quants. Both Ran @ 110-120tokens/sec in OpenCode 2+LM Studio. — retained record

> Retained reddit record from reddit:LocalLLaMA.

Canonical: https://highsignal.app/data/records/52d0534ae19b01e4

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


source: [reddit](https://highsignal.app/data/reddit) (reddit:LocalLLaMA)  published 2026-11-09 01:33 ingested 2026-11-10 02:42

[original source ↗](https://www.reddit.com/r/LocalLLaMA/comments/1xrexe0/quick_one_shot_pelican_rtx_5090_test_qwen_38_27b/)

## retained text



```
Test - A pelican riding a bicycle.  
GPU - RTX 5090 (31GB VRAM Usage, Model+KV cache)  
Speed - 110-120tok/sec  
Model - Qwen 3.8 27B Unsloth, GGUF  
Quants/Context - Q4\_K\_M/120K vs. Q5\_K\_M/90K  
Software - OpenCode 2+LM Studio(MTP On, Thinking Medium, All Else Default)
```



document fetched 2026-11-10 00:41 · canonical https://reddit.com/r/LocalLLaMA/comments/1xrexe0/quick_one_shot_pelican_rtx_5090_test_qwen_38_27b
