AI/TLDR

Two Minute Papers · 2026-08-05 · notable

Two Minute Papers — 'The Billion Dollar AI Race Just Broke'

Two Minute Papers walks through Alibaba's Qwen3.8-Max — a 2.4T-parameter MoE that clears the price-performance frontier days after launch. Károly Zsolnai-Fehér frames it as the latest jolt to the US-China AI race.

Two Minute Papers thumbnail — 'The Billion Dollar AI Race Just Broke'.

Two Minute Papers on Qwen3.8-Max: 'the billion dollar AI race just broke'.

What is it?

A fresh Two Minute Papers episode from Károly Zsolnai-Fehér that unpacks Alibaba's Qwen3.8-Max release. The video walks through the 2.4T-parameter MoE model, its benchmark showing, and why Chinese open-weights labs keep undercutting closed US frontier labs on price.

How does it work?

Zsolnai-Fehér pulls the numbers from Alibaba's release page, compares Qwen3.8-Max scores against Claude Fable 5 and GPT-5.6, and highlights the $2/$6 per-million-token API pricing plus the promised open-weight drop. Standard Two Minute Papers pace — dense visuals, three-minute run, one 'what a time to be alive'.

Why does it matter?

The Two Minute Papers audience of ~1.6M subscribers is where many practitioners first hear about new frontier models. The episode signals just how quickly Qwen3.8-Max became the talk of the AI research community and reinforces the week's 'China caught up' narrative.

Sources · 2 outlets

Tags

  • two-minute-papers
  • video
  • qwen
  • qwen-3-8-max
  • alibaba
  • open-weights
  • china-ai

← All releases · Learn AI