Cogito v2 Preview (Deep Cogito research blog) 31 Jul 2025
- MMLU (thinking) 91.3
- MMLU-Pro (thinking) 85.4
- GPQA - Diamond (thinking) 76
- GSM8K (thinking) 96.2
- MATH (thinking) 98.2
- MMMLU (thinking) 82.4
- MGSM (thinking) 93
- Simple QA (thinking) 31.4
Cogito v2 Preview 671B MoE
Cogito v2 Preview 671B MoE ranks first on 2 of the 8 benchmarks it has scores on.
8
Benchmarks with a score
2
Ranked first, out of five or more models
2
In the top three
Deep Cogito Jul 2025 Cogito v2 Preview 671B MoE
Open weights Flagship
No Epoch Capabilities Index score yet
- MMLU-Pro 85.4% #21 /151
- MMMLU 82.4% #36 /62
- MMLU 91.3% #4 /127
Where it ranks
Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.
Knowledge and language
Math and science
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| MATH | 98.2% | #1 of 93 | Model card | |
| GSM8K | 96.2% | #6 of 87 | Model card | |
| GPQA Diamond | 76% | #135 of 284 | Model card |
Its launch post
The benchmarks the lab chose to headline, with the scores it gave.