Qwen4 Max scores 43 on Artificial Analysis Intelligence Index, 9 points above Qwen3.8 Max
Artificial Analysis reports that Alibaba's Qwen4 Max scores 43 on its Intelligence Index using xhigh reasoning effort. That is a gain of 9 points over Qwen3.8 Max, and 3 points behind Kimi K3.1 Pro.
Alibaba's Qwen4 Max scores 43 on the Artificial Analysis Intelligence Index, a gain of 9 points over Qwen3.8 Max, according to a post from Artificial Analysis (@artificialanlys) on X.
Key results
- Score: Qwen4 Max, run with xhigh reasoning effort, scores 43 on the Intelligence Index.
- Versus Qwen3.8 Max: The model gains 9 points over its predecessor. Artificial Analysis puts the cost per task at roughly 67% lower than Qwen3.8 Max.
- Versus Kimi K3.1 Pro: Qwen4 Max lands 3 points behind Kimi K3.1 Pro, at roughly 11% lower cost per task.
Artificial Analysis states that Alibaba has launched Qwen4 Max. Context window, per-token pricing, parameter count and training cutoff are not given in the feed summary, so none are reported here.
The cost figures refer to cost per task as measured by Artificial Analysis, not list price per 1M tokens.
What this means
The results place Qwen4 Max ahead of Alibaba's previous Max model on both measures Artificial Analysis reports: a higher Intelligence Index score and a lower cost per task. Against Kimi K3.1 Pro, the comparison runs the other way on capability, with Kimi ahead by 3 points, while Qwen4 Max is cheaper per task by roughly 11%. For teams choosing between the two on this index, the trade-off is a small score gap against a modest cost difference. The scores were measured at xhigh reasoning effort, so they apply to that setting.
Related Articles
Gemini 4 Hydrogen scores 54 on Artificial Analysis Intelligence Index, just below Claude Sonnet 5.5 and GPT-6.1 Sol
Artificial Analysis reports that Google's Gemini 4 Hydrogen scores 54 on its Intelligence Index, just below Claude Sonnet 5.5 and GPT-6.1 Sol. It costs a third of Sonnet's cost but about 2.5x that of GPT-6.1 Sol, and adds an xhigh reasoning level.
Gemini 4 Carbon tops MathArena at 85.4%, ahead of Claude Fable 5.2 and GPT-6.1 Sol
Google's Gemini 4 Carbon (high) leads the MathArena math benchmark with an expected performance of 85.4%, ahead of Anthropic's Claude Fable 5.2 (max) at 82.8% and OpenAI's GPT-6.1 Sol (max) at 80.6%. With the reasoning settings used in testing, Carbon costs roughly 15 times as much as GPT-6.1 Sol.
Co-Design Bench: best coding agent fully solves only 26.4% of real user change requests
Researchers from Alibaba and several Asian universities released Co-Design Bench, a coding-agent test built from real user change requests. According to the paper, the best of ten models, GPT-6 Astra, meets about 70 percent of a task's requirements on average but is fully correct on only 26.4 percent of tasks.
ACE Music v1 Tops Artificial Analysis Vocal Music Board at 1160 Elo, Within Margin of Suno v6
ACE Music v1 ranks first on Artificial Analysis' AA-Music-Vocal v1.1 leaderboard with an Elo of 1160 (±21), ahead of Suno v6 at 1144 (±18). The 16-point gap falls inside overlapping 95% intervals, so the board does not separate the two with confidence.
Comments
Loading...