- Who
- Alibaba_Qwen
- What
- reported benchmark results Qwen4-Max’s Agent Arena performance compared with Qwen3.8-Max
- Quote
Qwen4-Max outperforms Qwen3.8-Max by +7.5% net on Agent Arena, making Alibaba #5 lab overall.
What happened
Qwen says Qwen4-Max outperforms Qwen3.8-Max by +7.5% net on Agent Arena, making Alibaba the #5 lab overall. The quoted Agent Arena post says Qwen4-Max debuted at #14 overall, up from Qwen3.8-Max at #39, based on 14.7k real-world agent sessions.
Why it matters
Qwen4-Max ranks ahead of Qwen3.8-Max on Agent Arena, with gains reported across all four signals and all three categories; the results draw on 14.7k real-world agent sessions.
Read next
See the full storyQwen4-Max Ranks Above Qwen3.8-Max on Agent ArenaQwen4-Max outperforms Qwen3.8-Max by +7.5% net on Agent Arena, making Alibaba #5 lab overall.
View the quoted post on XQwen4-Max by @Alibaba_Qwen is the #1 Chinese model on the leaderboard making Alibaba now the #1 Chinese lab and #5 lab overall, just behind Anthropic, Google, OpenAI and Meta.
It debuts overall at #14 on Agent Arena with +7.5% net improvement across 14.7k real-world agentic sessions from our global community, real traces on real tasks.
This is a major leap over Qwen3.8-Max
- Rank: #39 → #14 overallThe Chinese frontier on Agent Arena:
- #14 Qwen4-Max (Alibaba): +7.5%
- #19 Hy4 (Tencent): +5.7%
- #21 Kimi K3.1 Pro (Max) (Moonshot): +4.0%
- #26 MiMo-V2.6-Pro (Xiaomi): +1.5%
- #30 DeepSeek-V4.1-Flash (Max) (DeepSeek): +1.1%
Breaking down performance across signals:
Qwen4-Max improves over Qwen3.8-Max across all four reported signals, with particularly large gains in user satisfaction and steerability.
- #15 Praise vs. Complaint: +13% (best Chinese model)
- #13 Confirmed Success: +9% (best Chinese model)
- #19 Steerability: +8%
- #29 Bash Reliability: +3%
By category, it is a top-ranked Chinese model across all 3:
- #12 in Work: +9%
- #14 in Chat: +8%
- #16 in Code: +8%
Congrats to the @Alibaba_Qwen team on this release!
The story so far
Qwen says Qwen4-Max outperformed Qwen3.8-Max by 7.5% net on Agent Arena; the quoted leaderboard post placed Qwen4-Max at No. 14 overall and Qwen3.8-Max at No. 39, based on 14,700 real-world agent sessions. Qwen also said Alibaba ranked No. 5 among labs.
