benchmark
.
space
benchmarks
rankings
compare
voices
transcripts
articles
OpenAI
gpt-oss-20b
4 benchmarks
AIME
#4 of 39
98.7%
MMLU Pro
#10 of 29
85.3%
GPQA Diamond
#45 of 49
67.1%
SWE-bench Verified
#37 of 40
60.7%