Skip to content
tuan3w / qwen-warpView on GitHub
Fast Qwen3.8-27B for one RTX 3090. 200k context, 100+ tok/s. Custom CUDA decode engine, OpenAI-compatible API.
☆24Nov 9, 2026Updated this week

Alternatives and similar repositories for qwen-warp

Users that are interested in qwen-warp are comparing it to the libraries listed below. We may earn a commission when you buy through links labeled 'Ad' on this page.

Sorting:

Are these results useful?