Kimi K3.1 Pro (Max): benchmark results
Provider: Moonshot. Released 2026-10-23. Access: API.
Aggregate Rating 1963 ± 13, rank #5 of 2166 rated models, from 57 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| BrowseComp | 92.1 | Score (%) | 100 |
| Kimi K3.1 Pro Tech Blog - Agents' Last Exam (ALL) | 43.2 | Score (%) | 100 |
| Kimi K3.1 Pro Tech Blog - Agents' Last Exam (CLI) | 44.2 | Score (%) | 100 |
| Kimi K3.1 Pro Tech Blog - BenchCAD w/ tools | 92.7 | Voxel IoU (%) | 100 |
| Kimi K3.1 Pro Tech Blog - BenchCAD w/o tools | 72.8 | Voxel IoU (%) | 100 |
| Kimi K3.1 Pro Tech Blog - BioMysteryBench | 80.6 | Score (%) | 100 |
| Kimi K3.1 Pro Tech Blog - BrowseComp | 92.1 | Score (%) | 100 |
| Kimi K3.1 Pro Tech Blog - DECK-Arena | 1456 | Elo | 100 |
| Kimi K3.1 Pro Tech Blog - GDP.pdf | 34.6 | Strict response pass rate (%) | 100 |
| Kimi K3.1 Pro Tech Blog - GDP.pdf w/ python | 85.7 | Task-macro rubric pass rate (%) | 100 |
| Kimi K3.1 Pro Tech Blog - HANDBOOK.md | 44.6 | Score (%) | 100 |
| Kimi K3.1 Pro Tech Blog - HLE-Full w/ tools | 70 | Score (%) | 100 |
Interactive version: theaggregate.ai/model?slug=kimi-k3-1-pro-max · How It Works · Data refreshed daily, snapshot 2026-10-25.