ainewsblitz.com

Breaking

Alibaba open-weights Qwen3.8-Flash, a 125B MoE preview of Qwen4 priced for global use

  • Foundation Models
  • Open Source
  • AI Agents

Alibaba Group Holding Ltd. on Wednesday made Qwen3.8-Flash downloadable and published its weights, positioning a smaller, cheaper Qwen model as a global alternative to larger frontier systems. The company said the 125-billion-parameter Mixture-of-Experts release is competitive with Anthropic’s Claude Opus 4.6 and DeepSeek’s V4-Flash, and is meant to widen adoption of its flagship Qwen line rather than only chase peak scale.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.

$20
Read this article
$29/month
Unlimited — all 27,523 articles, the full archive, and comprehension quizzes
Save 72%
$98/year
≈ $8.17/month
Unlimited, billed once a year