Kimi
Kimi K3 is a 2.8T parameter open-weight multimodal reasoning model from Moonshot AI. It is suited for complex coding, knowledge work, and long-horizon agentic workflows, and is particularly strong at...
Kimi model releasesRank #40 across 600
Rank #19 across 266
Rank #337 across 379
Percentile score by analysis domain.
* Cost is inverted: lower input, output, and blended prices rank higher.
Higher bars mean stronger relative placement.
| Metric | Domain | Value | Rank |
|---|---|---|---|
| Artificial Analysis Intelligence Index | overall | 46.6 | #40 |
| Artificial Analysis Coding Index | coding | 72.0 | #19 |
| GPQA | reasoning | 84.2% | #110 |
| Humanity's Last Exam | reasoning | 24.0% | #96 |
| SciCode |
| coding, reasoning |
| 51.2% |
| #35 |
| Output Speed | speed | 35.9 tok/s | #158 |
| Time to First Token | speed | 2.48s | #115 |
| Blended Price | cost | $6.00/M | #337 |
| Input Price | cost | $3.00/M | #337 |
| Output Price | cost | $15.00/M | #339 |
| Value Index | cost, overall | 7.8 | #281 |