Qwen3.8-27B Hits 417 tok/s on a Single RTX 3090 With 150k Context

A community project squeezes Qwen3.8-27B onto a 24GB gaming card with vLLM, hitting 417 tok/s batched or 82 tok/s single-user at 150k context.

Source

AlphaSignal

Published

Aug 27, 2026, 8:51 AM

Author

AlphaSignal Newsroom

Read

1 min read

A community project squeezes Qwen3.8-27B onto a 24GB gaming card with vLLM, hitting 417 tok/s batched or 82 tok/s single-user at 150k context.

Reporting is indexed from AlphaSignal. Rights remain with the original publisher and cited sources.

Next reads