Model creator
Moonshot AI
Every Moonshot AI model in the BenchSift catalogue, with its scores, ranks, prices and capabilities as the sources publish them.
- Models
- 12
- 16 configurations, reasoning levels included
- Open weights
- 11
- 11 of 12 models
- Latest release
- Oct 23, 2026
- Kimi K3.1 Pro · First dated model: Jul 11, 2025
- Median price
- $1.71
- $1.00 to $6 per million tokens, across 7 priced models
Models
12Reasoning levels are folded into one entry. The rank is the whole catalogue's: sorting does not change it.
- Kimi K3.1 Pro46.5#33 of 782$432 t/s
Text · Reasoning · Open weights · Oct 23, 2026 · $4 · 32 t/s
- Kimi K2.627.0#161 of 782$1.7135.4 t/s
Text · Reasoning · Open weights · 2 configurations · Apr 20, 2026 · $1.71 · 35.4 t/s
- Kimi K2.7 Code25.8#175 of 782$1.7159.9 t/s
Text · Reasoning · Open weights · Jun 12, 2026 · $1.71 · 59.9 t/s
- Kimi K2.523.5#208 of 782$1.2043.1 t/s
Text · Reasoning · Open weights · 2 configurations · Jan 27, 2026 · $1.20 · 43.1 t/s
- Kimi K317.4#300 of 782$645.2 t/s
Text · Reasoning · Open weights · 2 configurations · Jul 16, 2026 · $6 · 45.2 t/s
- Kimi K2 090515.3#325 of 782$1.0843.4 t/s
Text · Open weights · Sep 5, 2025 · $1.08 · 43.4 t/s
- Kimi K212.7#374 of 782$1.0037.6 t/s
Text · Reasoning · Open weights · 2 configurations · Jul 11, 2025 · $1.00 · 37.6 t/s
- Kimi Linear 48B A3B Instruct7.3#575 of 782——
Text · Open weights · Oct 30, 2025
- Kimi K2.8 Preview———
Text · Reasoning · Sep 11, 2026
- Kimi K2.7 Code Highspeed———
Text · Reasoning · Open weights · Jun 12, 2026
- Kimi K2 Thinking Turbo———
Text · Reasoning · Open weights · Nov 6, 2025
- Kimi K2 Instruct———
Text · Open weights · Jul 11, 2025
Where these figures come from
Each catalogue model is tied to its creator by the identity Artificial Analysis publishes, or OpenRouter's and Models.dev's for models it does not track. Counts fold the reasoning levels of one model together. Ranks compare each measure with every measured model in the catalogue, configurations included, as model pages do. The price is the published blended price (3 input tokens to 1 output token).
Figures computed by BenchSift from the catalogue: Artificial Analysis, OpenRouter, Models.dev and Hugging Face. A value no model carries is not shown.
Data retrieved on Nov 12, 2026.