Deep Cogito: Cogito v2.1 671B
Deep Cogito's Cogito v2.1 671B has no intelligence index score, which limits its overall rating, but shows reasonable GPQA and MMLU Pro results alongside strong math performance — a mid-tier option for content and reasoning tasks from a less established provider, without benchmark data sufficient to confirm top-tier status.
Assessment date: March 12, 2026
Our methodology takes into account a range of factors including pricing, functionality, capabilities, benchmark performance, and real-world applicability. Rankings are reviewed and updated regularly as new models are released. Issues with our rankings? Contact us
Cogito v2.1 671B MoE represents one of the strongest open models globally, matching performance of frontier closed and open models. This model is trained using self play with reinforcement learning to reach state-of-the-art performance on multiple categories (instruction following, coding, longer queries and creative writing). This advanced system demonstrates significant progress toward scalable superintelligence through policy improvement.
Architecture
| Modality | Text → Text |
| Tokenizer | Other |
| Parameters | 671B |
Performance Indices
Source: Artificial Analysis
Benchmark Scores
Evaluations
Benchmark data from Artificial Analysis and Hugging Face
Model Information
Pricing
| Token Type | Cost per 1M tokens | Cost per 1K tokens |
|---|---|---|
| Input | $1.25 | $0.001250 |
| Output | $1.25 | $0.001250 |
Live Performance
Live endpoint metrics — refreshed every 30 minutes.
