Artificial Analysis: Qwen4 Max scores 43 on its Intelligence Index, 3 behind Kimi K3.1 Pro
Artificial Analysis says Qwen4 Max, at xhigh reasoning effort, scores 43 on its Intelligence Index: 9 points above Qwen3.8 Max (34) and 3 below Kimi K3.1 Pro (46).
- Gains: it improves on every evaluation in the index. HLE-Diamond rises from 24% to 45%, and AA-Omniscience, AutomationBench-AA, CritPt v2, Terminal-Bench-Science and Terminal-Bench 4.0 also rise.
- Cost: $3.27 per Intelligence Index task, about 3x less than Qwen3.8 Max ($10.02) and about 11% less than Kimi K3.1 Pro ($3.68). It uses about 105k output tokens per task, 39% fewer than Qwen3.8 Max, though more than Kimi K3.1 Pro (about 72k).
- Where it trails: Terminal-Bench 4.0 at 48%, against 67% for Kimi K3.1 Pro; FrontierCode 2.0 at 26%, level with Kimi K3.1 Pro and behind Tencent Hy4 (31%).
- Knowledge work: 1686 Elo on AA-Briefcase, against 1674 for Kimi K3.1 Pro.
Artificial Analysis lists a 1M-token context window, text and image input, and $1.50/$4.50 per million input/output tokens on Alibaba's own API.
Sources
Discussions
- So Qwen4 Max scores 43 on the AA intelligence index 😭X · 43 comments
- Qwen 4 is not a good release.X · 126 comments
Get news like this as a push.
AI Tracker alerts you when something changes for the tools you use.