2 vs 0 benchmarks won
Anthropic Claude Sonnet 4.6 | Moonshot AI Kimi K2 (0905) | |
|---|---|---|
| Overview | ||
| Company | Anthropic | Moonshot AI |
| Release date | Feb 17 2026 | Sep 5 2025 |
| Model type | — | — |
| Open source | No | Yes |
| Specifications | ||
Parameters | — | 1T |
Context window | — | 256k |
| Benchmarks | ||
Science reasoning GPQA Diamond | 89.9% | 75.1% |
Software engineering SWE-Bench Verified | 79.6% | 65.8% |
Multimodal understanding MMMU | — | — |
| Timeline | ||
| Release gap | Kimi K2 (0905) shipped 165 days before Claude Sonnet 4.6 | |
Claude Sonnet 4.6 leads Kimi K2 (0905) on 2 of the tracked benchmarks (GPQA Diamond, SWE-Bench Verified, MMMU). Kimi K2 (0905) shipped 165 days before Claude Sonnet 4.6, so benchmark comparisons should account for the intervening progress.
Kimi K2 (0905) is an open-source / open-weight model; Claude Sonnet 4.6 is proprietary.
On GPQA Diamond, Claude Sonnet 4.6 scores 89.9%, 14.8 points above Kimi K2 (0905) at 75.1%. On SWE-Bench Verified, Claude Sonnet 4.6 scores 79.6%, 13.8 points above Kimi K2 (0905) at 65.8%.