MAMoonshot AI/ Kimi
Kimi K3
moonshotai/kimi-k3
Multimodal Kimi model with 1M context and toggleable max-effort thinking for long-horizon agent work
Model information
Released
2026-07-16
Context window
1M
Max output
131K
Weights
Open
Modalities
Input
ImageTextVideo
Output
Text
Identity
- Version
- kimi-k3
- Aliases
- Kimi K3
Resources
Capabilities
AttachmentsOpen weightReasoningStructuredTool call
Official price
input $3/M · output $15/M
cache read $0.3/M
Providers
| Provider | Provider model ID | Protocol | Context | Max output | Input | Output | Capabilities | Notes | Pricing | Streaming |
|---|---|---|---|---|---|---|---|---|---|---|
| Weights & Biases | moonshotai/Kimi-K3 | openai_chat | 1M | 131K | AttachmentsReasoningTool callStructuredTemperature | - | input $3/M · output $15/M cache read $0.3/M | Yes | ||
| OpenCode Go | kimi-k3 | openai_chat | 1M | 131K | AttachmentsReasoningTool callStructured | - | input $3/M · output $15/M cache read $0.3/M | Yes | ||
| Baseten | moonshotai/Kimi-K3 | openai_chat | 1M | 131K | AttachmentsReasoningTool callStructuredTemperature | - | input $3/M · output $15/M | Yes | ||
| Nebius Token Factory | moonshotai/Kimi-K3 | openai_chat | 1M | 131K | ReasoningTool callStructured | - | input $3/M · output $15/M cache read $3/M | Yes | ||
| CrofAI | kimi-k3 | openai_chat | 1M | 131K | AttachmentsReasoningTool callStructured | - | input $2/M · output $8/M cache read $0.25/M | Yes | ||
| LLM Gateway | kimi-k3 | openai_chat | 1M | 131K | AttachmentsReasoningTool callStructured | - | input $3/M · output $15/M cache read $0.3/M | Yes | ||
| Kenari | kimi-k3 | openai_chat | 1M | 131K | AttachmentsReasoningTool callStructured | - | input $0/M · output $0/M | Yes | ||
| OpenCode Zen | kimi-k3 | openai_chat | 1M | 131K | AttachmentsReasoningTool callStructured | - | input $3/M · output $15/M cache read $0.3/M | Yes | ||
| ai& | moonshotai/kimi-k3 | openai_chat | 1M | 131K | AttachmentsReasoningTool callStructured | - | input $3/M · output $12.5/M cache read $0.5/M | Yes | ||
| ZenMux | moonshotai/kimi-k3 | openai_chat | 1M | 131K | AttachmentsReasoningTool callStructured | - | input $3/M · output $15/M cache read $0.3/M | Yes | ||
| EmpirioLabs AI | kimi-k3 | openai_chat | 1M | 131K | AttachmentsReasoningTool callStructured | - | input $3/M · output $15/M cache read $3/M | Yes | ||
| Venice AI | kimi-k3 | openai_chat | 1M | 131K | AttachmentsReasoningTool callStructured | - | input $3.75/M · output $18.75/M cache read $0.375/M | Yes | ||
| Together AI | moonshotai/Kimi-K3 | openai_chat | 1M | 131K | AttachmentsReasoningTool callStructured | - | input $3/M · output $15/M cache read $0.3/M | Yes | ||
| Moonshot AI (China) | kimi-k3 | custom | 1M | 131K | AttachmentsReasoningTool callStructured | - | input $3/M · output $15/M cache read $0.3/M | Yes | ||
| Modal | moonshotai/Kimi-K3 | openai_chat | 1M | 131K | AttachmentsReasoningTool callStructured | - | input $3/M · output $15/M cache read $0.3/M | Yes | ||
| Berget.AI | moonshotai/Kimi-K3 | openai_chat | 1M | 131K | AttachmentsReasoningTool callStructured | - | input $3/M · output $15/M | Yes | ||
| AIHubMix | kimi-k3 | openai_chat | 1M | 131K | AttachmentsReasoningTool callStructured | - | input $3/M · output $15/M cache read $0.3/M | Yes | ||
| Charm Hyper | kimi-k3 | openai_chat | 1M | 131K | AttachmentsReasoningTool callStructured | - | input $3.266/M · output $16.332/M cache read $0.327/M | Yes | ||
| Vercel AI Gateway | moonshotai/kimi-k3 | openai_chat | 1M | 131K | AttachmentsReasoningTool callStructured | - | input $3/M · output $15/M cache read $0.3/M | Yes | ||
| NovitaAI | moonshotai/kimi-k3 | openai_chat | 1M | 131K | AttachmentsReasoningTool callStructured | - | input $3/M · output $15/M cache read $0.3/M | Yes | ||
| OpenRouter | moonshotai/kimi-k3 | openai_chat | 1M | 131K | AttachmentsReasoningTool callStructuredTemperature | Reasoning levels: max, high, low, Default reasoning: max, Hugging Face: moonshotai/Kimi-K3 | input $3/M · output $15/M cache read $0.3/M | Yes | ||
| Hugging Face | moonshotai/Kimi-K3 | openai_chat | 1M | 131K | AttachmentsReasoningTool callStructured | - | input $3/M · output $15/M | Yes | ||
| Deep Infra | moonshotai/Kimi-K3 | openai_chat | 1M | 131K | AttachmentsReasoningTool callStructured | - | input $2.7/M · output $13.5/M cache read $0.27/M | Yes | ||
| Cloudflare AI Gateway | moonshotai/kimi-k3 | openai_chat | 1M | 131K | AttachmentsReasoningTool callStructured | - | input $3/M · output $15/M cache read $0.3/M | Yes | ||
| Ollama Cloud | kimi-k3 | openai_chat | 1M | 131K | AttachmentsReasoningTool callStructured | - | - | Yes | ||
| Moonshot AI | kimi-k3 | custom | 1M | 131K | AttachmentsReasoningTool callStructured | - | input $3/M · output $15/M cache read $0.3/M | Yes | ||
| Cortecs | kimi-k3 | openai_chat | 1M | 131K | AttachmentsReasoningTool callStructured | - | input $3/M · output $15/M | Yes | ||
| NanoGPT | moonshotai/kimi-k3 | openai_chat | 1M | 131K | AttachmentsReasoningTool callStructured | - | input $2.5/M · output $13.5/M cache read $0.25/M | Yes | ||
| TensorX | moonshotai/kimi-k3 | openai_chat | 1M | 131K | AttachmentsReasoningTool callStructured | - | input $3/M · output $15/M cache read $0.75/M | Yes | ||
| Kilo Gateway | moonshotai/kimi-k3 | openai_chat | 1M | 131K | AttachmentsReasoningTool callStructuredTemperature | - | input $3/M · output $15/M cache read $0.3/M | Yes |