Kimi series comparison
Moonshot's Kimi family: very long context, ranking among the top open-weight models.
All models
Kimi series comparison
Moonshot's Kimi family: very long context, ranking among the top open-weight models.
4 versions (sorted by capability)
| Version | Context | Input | Output | Reasoning | Code | Speed | Open source |
|---|---|---|---|---|---|---|---|
| Kimi K2 | 256K | — | — | ●●●●○ | ●●●●○ | ●●●●○ | Open |
| Kimi K2 Thinking | 256K | — | — | ●●●●● | ●●●●○ | ●●●○○ | Open |
| Kimi K2.5 | 256K | — | — | ●●●●○ | ●●●●○ | ●●●●○ | Open |
| Kimi K2.6 | — | — | — | ●●●●○ | ●●●●○ | ●●●●○ | Open |
Kimi K2Kimi K2: 128K–256K context, 1T total params (32B active) MoE, trained on 15.5T tokens, open-source weights.Kimi K2 ThinkingKimi K2 Thinking: 256K context, 1T parameters (32B activated) MoE, focused on deep reasoning, open-source weights.Kimi K2.5Moonshot Kimi K2.5: 256K context, 1T total params (32B active) MoE + vision, open-source weights, multimodal upgrade of K2.Kimi K2.6Moonshot Kimi series, one of the highest-ranked open-source weight models on the current Intelligence Index.