XNews

Alibaba releases Qwen3‑Next‑80B FP8 models with Transformers and vLLM compatibility

Alibaba announced Qwen3‑Next‑80B‑A3B‑Instruct‑FP8 and Qwen3‑Next‑80B‑A3B‑Thinking‑FP8, now fully compatible with Transformers, vLLM and SGLang, and using FP8 precision to enable faster inference. Downloads available on Hugging Face and ModelScope.

Source posts