Qwen4-Max scores 43 on the Artificial Analysis Intelligence Index, beating Qwen3.8 Max at ~67% lower cost per task but trailing Kimi K3.1 Pro at ~11% lower cost (@artificialanlys)
Qwen4-Max uses 39% fewer output tokens per benchmark task than its predecessor.
Artificial Analysis reports roughly 105,000 output tokens per task, down from 171,000 for Qwen3.8 Max. Its Intelligence Index score rises nine points to 43, while cost per task falls from $10.02 to $3.27. Kimi K3.1 Pro scores three points higher at $3.68 per task. Tencent Hy4 costs $1.03 per task but scores eight points below Qwen4-Max.
Techmeme's note
Artificial Analysis reports roughly 105,000 output tokens per task, down from 171,000 for Qwen3.8 Max. Its Intelligence Index score rises nine points to 43, while cost per task falls from $10.02 to $3.27. Kimi K3.1 Pro scores three points higher at $3.68 per task. Tencent Hy4 costs $1.03 per task but scores eight points below Qwen4-Max.
Techmeme's note
score 7