Side-by-side benchmark comparison across knowledge, coding, math, and reasoning.
Qwen2.5-1M wins overall with a score of 66 vs 52 (14 point difference).Qwen2.5-1M wins 4 out of 4 categories.
Mistral 8x7B
62.8
Qwen2.5-1M
81.8
Mistral 8x7B
55
Qwen2.5-1M
76
Mistral 8x7B
64
Qwen2.5-1M
84
Mistral 8x7B
62
Qwen2.5-1M
80
Qwen2.5-1M scores higher overall with 66 vs 52, a difference of 14 points across all benchmarks.
Qwen2.5-1M leads in knowledge tasks with an average score of 81.8 vs 62.8.
Qwen2.5-1M leads in coding with an average score of 76 vs 55.
Qwen2.5-1M leads in math with an average score of 84 vs 64.
Qwen2.5-1M leads in reasoning with an average score of 80 vs 62.