Qwen3.8-Max Released: 2.4T Parameters, Competitive Pricing, and Surprising Writing Skills
数字生命卡兹克 · wechat · 2026-08-03
Alibaba's Qwen3.8-Max has been officially released, featuring 2.4T total parameters, 95B activated parameters, and a 1M context window. The model is set to be open-sourced next week, alongside a smaller 27B version for local deployment.
Hands-on testing reveals its overall performance is on par with Kimi K3, but it boasts significant pricing advantages ($2/M input, $6/M output) and faster inference speeds. The author specifically highlights that, unlike competitors whose writing capabilities have regressed due to over-optimization for coding, Qwen3.8-Max retains an impressive sense of language and writing quality, making it a highly competitive all-around model.
Related event: Alibaba releases 2.4T-parameter Qwen3.8-Max, open-sourcing next week(26 posts)→
More from Models
- Qwen 3.8 Max Outperforms Fable 5 in 3D Physics Scene Generation at 1/7 the Cost — rohanpaul_ai · 2026-08-03
- OpenAI's New Astra Model, AI Agents Escaping Sandboxes, and Pacing Calls — EverydayAI_ · 2026-08-03
- llama.cpp Adds MTP Support for Qwen3-Next, Enabling Full-Speed Inference — jacek2023 · 2026-08-03
- Chinese Labs Moonshot and Alibaba Crack Code Arena's Top 5 WebDev Leaderboard — heypearlai · 2026-08-03
- Qwen 3.8 Max Released, Praised as Cost-Effective; Cola App Lauds Its Capabilities — oran_ge · 2026-08-03
- DeepSeek v4 Flash Sets New Cost/Accuracy SOTA on WeirdML — teortaxesTex · 2026-08-03