Alibaba Announces Qwen3.8-Max (2.4T) with 27B Open-Weights Coming Next Week
victormustar · x · 2026-08-03
Alibaba has announced Qwen3.8-Max, its most capable model to date at 2.4T parameters. The open weights for the Qwen3.8-27B version will be released next week. The company highlights a new bar for autonomous coding, claiming the model can evolve from an empty folder to production-grade code without hand-holding. It also features long-horizon mastery for system-level planning, handling hundreds of turns in chip design and e-commerce strategies.
Related event: Alibaba releases 2.4T-parameter Qwen3.8-Max, open-sourcing next week(23 posts)→
More from Models
- China's AI Strategy: Undercutting US Closed Models with Open-Weights — wschroll · 2026-08-03
- Qwen 3.8 Max and MiniMax-H3 Released Within Hours of Each Other — AccBalanced · 2026-08-03
- SupraBrain-50M released: fuses linear recurrence with attention — LH-Tech_AI · 2026-08-03
- Users Report Severely Reduced Claude Chat Limits Recently — Spartaness · 2026-08-03
- Big Tech's GPU Hoarding Raises Open-Weight Hosting Costs Near Closed-Model Levels — xiaosun86 · 2026-08-03
- DeepSeek V4 Flash Burns 8T Tokens Daily, Reshaping Agent-Era Model Pricing — APPSO · 2026-08-03