Qwen3.7 Preview Model Now Available: Enhanced Math, Coding, and Multimodal Capabilities
The rapid evolution of large language models continues to push boundaries. Alibaba has quietly introduced two preview iterations of its Tongyi Qwen 3.7 series—Qwen3.7-Max-Preview and Qwen3.7-Plus-Preview—on Qwen Chat and Arena AI (formerly LMArena). These releases are expected to be officially unveiled at the upcoming Alibaba Cloud Summit on May 20th.
These new previews serve distinct purposes:
Qwen3.7-Max-Preview acts as a flagship model focused on top-tier comprehensive performance. Currently, it operates in thinking mode only, with advanced tools like search and code interpretation yet to be enabled.
Qwen3.7-Plus-Preview offers a balanced high-performance experience, prioritizing reasoning and logical clarity. Its full suite of tools will be gradually rolled out in future updates.

Despite being in preview, these models have already garnered significant attention. On Arena AI’s text capability leaderboard, Qwen3.7-Max-Preview ranks 13th overall, propelling Alibaba’s Tongyi Qwen to sixth place globally among comparable labs. In specialized areas, its performance is even more striking: it places 7th in mathematical reasoning, 9th in expert-level applications, 9th in software and IT domains, and 10th in programming proficiency, showcasing robust multi-task generalization.
Regarding visual capabilities, Qwen3.7-Plus-Preview ranks 16th overall, helping Tongyi Qwen secure fifth place among all labs in visual benchmarks. Additionally, in the exclusive "Expert Arena" for advanced prompts, Qwen3.7-Max-Preview achieved 9th place, highlighting its growing stability and precision in high-difficulty tasks.
Related article
Meta Tests AI Bedtime Story App StoryKit to Solve Parents’ Storytelling Challenges
Designed to support busy parents, Meta is testing StoryKit, an AI-powered app that generates personalized bedtime stories for kids, complete with custom characters, scenes, and music.Parents simply upload photos of their child’s favorite toys to crea
OpenAI Launches GPT-5.3 Instant, Cutting Hallucinations by 27 Percent
Merely two hours after Google DeepMind unveiled Gemini 3.1 Flash-Lite, OpenAI moved with unprecedented velocity, officially releasing its most refined conversational model to date: GPT-5.3Instant.Unlike previous iterations that prioritized achieving
Zhipu Acquires Zhongke Jiahe in Hundreds-of-Millions-Yuan Deal to Boost AI Infrastructure and Domestic Computing Compatibility
Zhipu has reportedly invested hundreds of millions of RMB to acquire AI infrastructure firm XCore Sigma, bolstering its foundational model engineering, domestic computing power compatibility, and inference optimization.Specializing in AI heterogeneou
Related Special Topic Recommendations
Comments (0)
0/500
The rapid evolution of large language models continues to push boundaries. Alibaba has quietly introduced two preview iterations of its Tongyi Qwen 3.7 series—Qwen3.7-Max-Preview and Qwen3.7-Plus-Preview—on Qwen Chat and Arena AI (formerly LMArena). These releases are expected to be officially unveiled at the upcoming Alibaba Cloud Summit on May 20th.
These new previews serve distinct purposes:
Qwen3.7-Max-Preview acts as a flagship model focused on top-tier comprehensive performance. Currently, it operates in thinking mode only, with advanced tools like search and code interpretation yet to be enabled.
Qwen3.7-Plus-Preview offers a balanced high-performance experience, prioritizing reasoning and logical clarity. Its full suite of tools will be gradually rolled out in future updates.

Despite being in preview, these models have already garnered significant attention. On Arena AI’s text capability leaderboard, Qwen3.7-Max-Preview ranks 13th overall, propelling Alibaba’s Tongyi Qwen to sixth place globally among comparable labs. In specialized areas, its performance is even more striking: it places 7th in mathematical reasoning, 9th in expert-level applications, 9th in software and IT domains, and 10th in programming proficiency, showcasing robust multi-task generalization.
Regarding visual capabilities, Qwen3.7-Plus-Preview ranks 16th overall, helping Tongyi Qwen secure fifth place among all labs in visual benchmarks. Additionally, in the exclusive "Expert Arena" for advanced prompts, Qwen3.7-Max-Preview achieved 9th place, highlighting its growing stability and precision in high-difficulty tasks.
Meta Tests AI Bedtime Story App StoryKit to Solve Parents’ Storytelling Challenges
Designed to support busy parents, Meta is testing StoryKit, an AI-powered app that generates personalized bedtime stories for kids, complete with custom characters, scenes, and music.Parents simply upload photos of their child’s favorite toys to crea
OpenAI Launches GPT-5.3 Instant, Cutting Hallucinations by 27 Percent
Merely two hours after Google DeepMind unveiled Gemini 3.1 Flash-Lite, OpenAI moved with unprecedented velocity, officially releasing its most refined conversational model to date: GPT-5.3Instant.Unlike previous iterations that prioritized achieving
Zhipu Acquires Zhongke Jiahe in Hundreds-of-Millions-Yuan Deal to Boost AI Infrastructure and Domestic Computing Compatibility
Zhipu has reportedly invested hundreds of millions of RMB to acquire AI infrastructure firm XCore Sigma, bolstering its foundational model engineering, domestic computing power compatibility, and inference optimization.Specializing in AI heterogeneou





Home






