reasoning
BullshitBench v2
- #195medium of 1962.0%
Leader: Opus 4.8 · 95.0%
Mistral
Measured on 1 benchmarks.
General frontier
25.5
#92 of 136 models · best at medium
includes estimates: 5% of the mix is measured for this model
reasoning
Leader: Opus 4.8 · 95.0%