Qwen3.8 27B Uncensored

on 2× NVIDIA RTX 5090 · vLLM · 262,144 ctx

Sep 27, 2026

Throughput

81.3 t/s gen · 3160.0 tokens/s prompt processing (prefill)Prompt processing measures input prefill speed; it does not tell you the time to first token.

Quant

FP8 (FP8)

KV cache

fp8

System RAM

32 GB

VRAM reported

48 GB

Summary

User reports Qwen3.8-27B Uncensored at 81.3 tok/s generation on an RTX 5090 plus RTX 4070 Ti Super. Setup is vLLM 0.30.0 with block-FP8 weights, fp8 KV cache, full 262,144 context, MTP K=3, and pipeline parallel across both cards. Prefill is 3,160 tok/s at 32K, 68.5-68.8 s for 145K fresh prefill, and 169 s for 258K prefill; MTP acceptance is about 87% and the 5090 peaks at 31,472 MiB.

View source ↗