Side-by-side benchmark comparison across knowledge, coding, math, and reasoning.
Qwen3 235B 2507 wins overall with a score of 30 vs 25 (5 point difference).Qwen3 235B 2507 wins 4 out of 4 categories.
DeepSeek V3.1 (Reasoning)
31.8
Qwen3 235B 2507
36.8
DeepSeek V3.1 (Reasoning)
26
Qwen3 235B 2507
31
DeepSeek V3.1 (Reasoning)
33
Qwen3 235B 2507
38
DeepSeek V3.1 (Reasoning)
31
Qwen3 235B 2507
36
Qwen3 235B 2507 scores higher overall with 30 vs 25, a difference of 5 points across all benchmarks.
Qwen3 235B 2507 leads in knowledge tasks with an average score of 36.8 vs 31.8.
Qwen3 235B 2507 leads in coding with an average score of 31 vs 26.
Qwen3 235B 2507 leads in math with an average score of 38 vs 33.
Qwen3 235B 2507 leads in reasoning with an average score of 36 vs 31.