Side-by-side benchmark comparison across knowledge, coding, math, and reasoning.
GPT-OSS 20B wins overall with a score of 22 vs 20 (2 point difference).GPT-OSS 20B wins 4 out of 4 categories.
GPT-OSS 20B
28.8
Mistral 8x7B v0.2
26.8
GPT-OSS 20B
23
Mistral 8x7B v0.2
21
GPT-OSS 20B
30
Mistral 8x7B v0.2
28
GPT-OSS 20B
28
Mistral 8x7B v0.2
26
GPT-OSS 20B scores higher overall with 22 vs 20, a difference of 2 points across all benchmarks.
GPT-OSS 20B leads in knowledge tasks with an average score of 28.8 vs 26.8.
GPT-OSS 20B leads in coding with an average score of 23 vs 21.
GPT-OSS 20B leads in math with an average score of 30 vs 28.
GPT-OSS 20B leads in reasoning with an average score of 28 vs 26.