modelgrep

Mistral: Mistral Medium 3.5 vs Qwen: Qwen3 Max Thinking

Qwen: Qwen3 Max Thinking wins on more metrics (4 of 7), but the right pick depends on what you optimize for — see the breakdown below.

MetricMistral: Mistral Medium 3.5Qwen: Qwen3 Max Thinking
Intelligence Index39.239.8
Coding Index35.430.5
GPQA Diamond75%86%
Design Arena Elo
Speed (tokens/sec)4738
Latency691ms1.3s
Input price /M$1.50$0.780
Output price /M$7.50$3.90
Context window262K262K
CapabilitiesReasoningToolsJSONVisionReasoningToolsJSON