Side-by-side benchmark comparison across knowledge, coding, math, and reasoning.
Qwen3.5 397B wins overall with a score of 65 vs 37 (28 point difference).Qwen3.5 397B wins 4 out of 4 categories.
Llama 4 Maverick
43.8
Qwen3.5 397B
80.8
Llama 4 Maverick
38
Qwen3.5 397B
75
Llama 4 Maverick
45
Qwen3.5 397B
82
Llama 4 Maverick
43
Qwen3.5 397B
79
Qwen3.5 397B scores higher overall with 65 vs 37, a difference of 28 points across all benchmarks.
Qwen3.5 397B leads in knowledge tasks with an average score of 80.8 vs 43.8.
Qwen3.5 397B leads in coding with an average score of 75 vs 38.
Qwen3.5 397B leads in math with an average score of 82 vs 45.
Qwen3.5 397B leads in reasoning with an average score of 79 vs 43.