Deep Cogito · Cogito
Cogito v2.1 (Reasoning)
Deep Cogito's 671B/37B-active open-weights reasoning model; Artificial Analysis publishes no Intelligence Index for it.
- AA Intelligence
- —
- Input / output per 1M
- $1.25/ $1.25
- Output speed
- —
- Context window
- 128K tokens
Coding
- LiveCodeBench#27 of 5668.8%
Leader: Fugu at 92.9%
Reasoning and knowledge
- GPQA Diamond#80 of 14276.8%
Leader: GPT-6 Astra at 96.1%
- Humanity’s Last Exam#89 of 14712%
Leader: Claude Opus 5.5 at 61.4%
- CritPt#137 of 1390%
Leader: GPT-5.6 Sol at 32.3%
Vision and long context
- AA Long Context Reasoning#111 of 13922.7%
Leader: Kimi K3 at 88.7%
Other published results
| Result | Score | Reported by | Source |
|---|---|---|---|
| aime-2025 · AIME 2025; reasoning | 72.7% | independent | Artificial Analysis: Cogito v2.1 (Reasoning) ↗ |
| Terminal-Bench Hard (AA) · reasoning | 16.7% | independent | Artificial Analysis: Cogito v2.1 (Reasoning) ↗ |
| IFBench · reasoning | 46.3% | independent | Artificial Analysis: Cogito v2.1 (Reasoning) ↗ |
| AA-Omniscience Index · reasoning | -25.8 index | independent | Artificial Analysis: Cogito v2.1 (Reasoning) ↗ |