Moonshot AIKimi
Kimi K3
Moonshot AI · Kimi. Released Jul 17, 2026.
Kimi K3 is a model from Moonshot AI in the Kimi family, released Jul 17, 2026. evals.report tracks 4 reported Kimi K3 benchmark scores across DeepSWE, Terminal-Bench 2.1, SWE-Marathon, BrowseComp — each shown with its benchmark, metric, source status, and date, and never combined into a single ranking.
Open4 results
Benchmark results 4
Compare this model| Benchmark | Category | Score | Metric | Status | Date | |
|---|---|---|---|---|---|---|
| DeepSWE | Coding | 67.5% | % resolved | Verified | Jul 17, 2026 | Details |
| Terminal-Bench 2.1 | Agents | 88.3% | task success | Verified | Jul 17, 2026 | Details |
| SWE-Marathon | Agents | 42.0% | resolution rate (pass@1) | Verified | Jul 17, 2026 | Details |
| BrowseComp | Agents | 91.2% | accuracy | Verified | Jul 17, 2026 | Details |