2 vs 0 benchmarks won
Anthropic Claude Opus 4.7 | Moonshot AI Kimi K2.5 | |
|---|---|---|
| Overview | ||
| Company | Anthropic | Moonshot AI |
| Release date | Apr 16 2026 | Jan 27 2026 |
| Model type | — | — |
| Open source | No | Yes |
| Specifications | ||
Parameters | — | 1T |
Context window | — | 256k |
| Benchmarks | ||
Science reasoning GPQA Diamond | 94.2% | 87.6% |
Software engineering SWE-Bench Verified | 87.6% | 76.8% |
Multimodal understanding MMMU | — | — |
| Timeline | ||
| Release gap | Kimi K2.5 shipped 79 days before Claude Opus 4.7 | |
Claude Opus 4.7 leads Kimi K2.5 on 2 of the tracked benchmarks (GPQA Diamond, SWE-Bench Verified, MMMU). Kimi K2.5 shipped 79 days before Claude Opus 4.7, so benchmark comparisons should account for the intervening progress.
Kimi K2.5 is an open-source / open-weight model; Claude Opus 4.7 is proprietary.
On GPQA Diamond, Claude Opus 4.7 scores 94.2%, 6.6 points above Kimi K2.5 at 87.6%. On SWE-Bench Verified, Claude Opus 4.7 scores 87.6%, 10.8 points above Kimi K2.5 at 76.8%.