Alibaba launched Qwen3.8-Max on Monday with its first published benchmarks, and the model ranked fourth on Arena’s Frontend Code leaderboard. Only Claude Opus 5 and Moonshot’s Kimi K3 scored higher in the test.
The 2.4 trillion-parameter model exits its July preview with 95 billion active parameters, a 1 million-token context window, and public API pricing.
Qwen3.8-Max Trails Only Claude and Kimi in Coding Test
Arena’s leaderboard scored Qwen3.8-Max at 1,668 points, one point behind Claude Opus 5 (High). Claude Opus 5 (Max) leads at 1,705, while Kimi K3 (Max) holds second at 1,676. The model also placed second in Arena’s Consumer Product category.
Follow us on X to get the latest news as it happens
Data shared by Alibaba claims wins over Claude Opus 4.8, Fable 5, and GPT-5.6 on agentic and multimodal benchmarks. The data shows 86.6 on TerminalBench-2.1, 93.0 on PaperBench, and 86.1 on OSWorld-Verified.
However, Anthropic’s Fable 5 keeps a wide lead in core software engineering. It scores 80.0 on SWE-Pro against Qwen’s 67.7, and 88.8 on FrontierSWE against 73.5.
First Open Max Model Arrives as Alibaba Shares Rally
The launch also marks the first time Alibaba will open-source the weights of a Qwen-Max-class model.
“Next week, the open weights of Qwen3.8-Max will be released, and Qwen3.8-27B is also going open-weights to meet you all!” the team said.
The weights will land on Hugging Face and ModelScope. Meanwhile, API pricing debuted at $2 per million input tokens and $6 per million output tokens.
Investors responded quickly to the launch. Alibaba shares climbed 6.15% to HK$124.20 by the Hong Kong lunch break, from a previous close of HK$117.
The stock extended Friday’s 4.65% gain, which analysts tied to a reported Moonshot chip deal. Independent evaluations may sharpen the picture once the weights land next week.
Subscribe to our YouTube channel to watch leaders and journalists provide expert insights
Source: BeInCrypto
