Side-by-side benchmark comparison across knowledge, coding, math, and reasoning.
Qwen3.5 397B wins overall with a score of 65 vs 39 (26 point difference).Qwen3.5 397B wins 4 out of 4 categories.
Llama 4 Behemoth
45.8
Qwen3.5 397B
80.8
Llama 4 Behemoth
40
Qwen3.5 397B
75
Llama 4 Behemoth
47
Qwen3.5 397B
82
Llama 4 Behemoth
45
Qwen3.5 397B
79
Qwen3.5 397B scores higher overall with 65 vs 39, a difference of 26 points across all benchmarks.
Qwen3.5 397B leads in knowledge tasks with an average score of 80.8 vs 45.8.
Qwen3.5 397B leads in coding with an average score of 75 vs 40.
Qwen3.5 397B leads in math with an average score of 82 vs 47.
Qwen3.5 397B leads in reasoning with an average score of 79 vs 45.