Victor M
@victormustar
๐ค Head of Product @huggingface
๊ฐ์
August 2012
2.3K ํ๋ก์ ์ค 30K ํฌ
here we go again: deployed a FREE public endpoint for Qwen3.8-Flash-Next ๐ (going at +100 tok/s)
No token needed, OpenAI-compatible, vision + tool calls, 262K context, thinking from xhigh โ off. Light rate limiting, be nice to your neighbors ๐ค
4ร H200 ยท FP8 ยท SGLang cookbook ยท ~140 tok/s per stream ยท ~100 tok/s @ 16 concurrent ยท 0.8s TTFT
Guide + chat UI ๐
๋ ๋ณด๊ธฐ
