Moonshot AI
4 ranked models in a catalog of 8 text models. Kimi K3 leads this lab at #9 overall.
Ranked models
Global ranks · current snapshot| Rank | Model | SI Score | coding | math | reasoning | preference | Confidence | Blended price | Context | Released |
|---|---|---|---|---|---|---|---|---|---|---|
| #9 | Kimi K3Open weights | 71.0 | 71.1 | 74.5 | 81.3 | 79.9 | 100% | $6.00 | 1M | Jul 16, 2026 |
| #32 | Kimi K2 Thinking TurboOpen weights | 66.1 | — | 83.1 | 84.2 | 74.5 | 64% | — | 262K | Nov 6, 2025 |
| #38 | Kimi K2.6Open weights | 64.0 | 62.6 | 70.6 | 77.5 | 78.2 | 85% | $1.71 | 262K | Apr 21, 2026 |
| #66 | Kimi K2.5Open weights | 59.0 | 72.4 | — | 37.3 | 77.3 | 100% | $1.20 | 262K | Jan 2026 |
Pillars use a 0–100 scale. * marks an imputed neutral prior where the model has no scored results in that pillar; it is not a measured benchmark result. Prices are USD per 1M tokens; see the blend and price comparison.
Scores and releases
Provisional models
4 awaiting rank-eligible evidence- Kimi K2.8 Preview
No usable open benchmark results found
- Kimi K2.7 Code
Evidence completeness 48.0%; at least 50% required
- Kimi K2.7 Code Highspeed
No usable open benchmark results found
- Kimi K2 Thinking
Results cover 1 pillar; at least 2 required