Code & Chain · Signal Desk

Alibaba Releases Qwen 3.8 Preview, a 2.4 Trillion Parameter Multimodal Model, Claiming Performance Second Only to Claude Fable 5

Original sourceOfficeChaiAdditional: MLQ News

Summary

Alibaba has released Qwen 3.8-Max-Preview, a 2.4 trillion parameter sparse mixture-of-experts multimodal model supporting text, image, video, and document inputs. In internal benchmarks, it ranks only behind Anthropic's Claude Fable 5, though independent third-party verification is pending. The mod…

Key points

  • Delivers near-frontier multimodal capabilities at lower cost and with API compatibility, reducing switching barriers for developers.
  • If weights are open-sourced, this would become the largest open-source multimodal model, potentially altering AI development cost structures.
  • Developers can immediately integrate via API and test multimodal applications at low cost, with the option for local deployment once weights are released.

Editorial note

This page is Code & Chain's editorial summary of public sources. It may be prepared with AI assistance and published through an automated workflow. Refer to the original sources; this content is not investment, legal, or tax advice.