ModelsSep 1, 2026 · 13:142 sources

Alibaba Releases Qwen3.8-Flash, First Open Weight Preview of Qwen4 Architecture

From huggingnews.com

Image: huggingnews.com
AI;DR

Alibaba has released Qwen3.8-Flash-Next, an open-weight preview of the Qwen4 architecture that uses a mixture-of-experts design with 125B parameters. The production version will cost $0.16 per 1 million input tokens and $0.47 per 1 million output tokens via QwenCloud API. The model supports a 262K native context window and ranks No. 7 among open models in the Agent Arena.

Read the sourceSign in to suggest

Sources

+ 1 more
  1. SUPPORT@thenewstack“An early preview of the architecture in Qwen4”x.com

One ranked edition, every morning

The stories that matter today. Free, cancel anytime.

Telegram @aidr_today