Moonshot · moonshotai/kimi-k3
kimi-k3
Senior Developer 1049k context Hermes StrongKimi 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...
Input / M
$3.000
Output / M
$15.000
Value score
20.37
30d price Δ
0%
Reasoning (AA)
43.6
Coding (AA)
76.2
Tool / agentic
—
Knowledge cutoff
—
Price history
Blended in+out list price, weekly (OpenRouterList) plus live snapshots. 13 points.
Capability snapshot
Last-known Artificial Analysis indexes via OpenRouter.
| Index | Score |
|---|---|
| coding | 76.2 |
| context | 100.0 |
| general_intelligence | 43.6 |
| reasoning | 43.6 |
| tool_use | 50.0 |