MAMoonshot AI/ Kimi
Kimi K2.5
moonshotai/kimi-k2.5
Earlier Kimi frontier model for long-context agents, coding, and multimodal work
Model information
Released
2026-01
Context window
262K
Max output
262K
Model parameters
1.1T
Weights
Open
Modalities
Input
ImageTextVideo
Output
Text
Identity
- Version
- kimi-k2.5
- Knowledge cutoff
- 2025-01
- Aliases
- Kimi K2.5
Resources
Capabilities
AttachmentsOpen weightReasoningStructuredTool call
Official price
input $0.6/M · output $3/M
cache read $0.1/M
Providers
| Provider | Provider model ID | Protocol | Context | Max output | Input | Output | Capabilities | Notes | Pricing | Streaming |
|---|---|---|---|---|---|---|---|---|---|---|
| Weights & Biases | moonshotai/Kimi-K2.5 | openai_chat | 262K | 262K | AttachmentsReasoningTool callStructuredTemperature | - | input $0.6/M · output $3/M cache read $0.1/M | Yes | ||
| OpenCode Go | kimi-k2.5 | openai_chat | 262K | 262K | AttachmentsReasoningTool callTemperature | - | input $0.6/M · output $3/M cache read $0.1/M | Yes | ||
| Baseten | moonshotai/Kimi-K2.5 | openai_chat | 262K | 262K | AttachmentsReasoningTool callStructuredTemperature | - | input $0.6/M · output $3/M cache read $0.12/M | Yes | ||
| Nebius Token Factory | moonshotai/Kimi-K2.5 | openai_chat | 262K | 262K | AttachmentsReasoningTool callStructuredTemperature | - | input $0.5/M · output $2.5/M cache read $0.05/M · cache write $0.625/M | Yes | ||
| NanoGPT | moonshotai/kimi-k2.5 | openai_chat | 262K | 262K | AttachmentsTool call | - | input $0.3/M · output $1.9/M cache read $0.15/M | Yes | ||
| DaoXE | kimi-k2.5 | openai_chat | 262K | 262K | AttachmentsReasoningTool callStructured | - | input $0.6/M · output $3/M cache read $0.1/M | Yes | ||
| CrofAI | kimi-k2.5 | openai_chat | 262K | 262K | AttachmentsReasoningTool callStructured | - | input $0.35/M · output $1.7/M cache read $0.07/M | Yes | ||
| Alibaba Coding Plan (China) | kimi-k2.5 | openai_chat | 262K | 262K | AttachmentsReasoningTool callTemperature | - | input $0/M · output $0/M cache read $0/M · cache write $0/M | Yes | ||
| LLM Gateway | kimi-k2.5 | openai_chat | 262K | 262K | AttachmentsReasoningTool callStructured | - | input $0.405/M · output $1.98/M cache read $0.225/M | Yes | ||
| OpenCode Zen | kimi-k2.5 | openai_chat | 262K | 262K | AttachmentsReasoningTool callTemperature | - | input $0.6/M · output $3/M cache read $0.08/M | Yes | ||
| Alibaba Coding Plan | kimi-k2.5 | openai_chat | 262K | 262K | AttachmentsReasoningTool callTemperature | - | input $0/M · output $0/M cache read $0/M · cache write $0/M | Yes | ||
| FrogBot | kimi-k2.5 | openai_chat | 262K | 262K | ReasoningTool callStructuredTemperature | - | input $0.6/M · output $3/M cache read $0.1/M | Yes | ||
| SiliconFlow | moonshotai/Kimi-K2.5 | openai_chat | 262K | 262K | ReasoningTool callStructuredTemperature | - | input $0.45/M · output $2.25/M cache read $0.07/M | Yes | ||
| ZenMux | moonshotai/kimi-k2.5 | openai_chat | 262K | 262K | AttachmentsReasoningTool call | - | input $0.58/M · output $3.02/M cache read $0.1/M | Yes | ||
| Abacus | kimi-k2.5 | openai_chat | 262K | 262K | ReasoningTool callStructuredTemperature | - | input $0.6/M · output $3/M | Yes | ||
| Alibaba Token Plan | kimi-k2.5 | openai_chat | 262K | 262K | AttachmentsReasoningTool callStructuredTemperature | - | input $0/M · output $0/M cache read $0/M · cache write $0/M | Yes | ||
| Azure Cognitive Services | kimi-k2.5 | custom | 262K | 262K | AttachmentsReasoningTool callStructuredTemperature | Interleaved output | input $0.6/M · output $3/M | Yes | ||
| Venice AI | kimi-k2-5 | openai_chat | 262K | 262K | AttachmentsReasoningTool callStructured | - | input $0.56/M · output $3.5/M cache read $0.22/M | Yes | ||
| Together AI | moonshotai/Kimi-K2.5 | openai_chat | 262K | 262K | ReasoningTool callTemperature | Interleaved output | input $0.5/M · output $2.8/M | Yes | ||
| DigitalOcean | kimi-k2.5 | openai_chat | 262K | 262K | AttachmentsReasoningTool callStructured | - | input $0.375/M · output $2.025/M cache read $0.203/M | Yes | ||
| Moonshot AI (China) | kimi-k2.5 | custom | 262K | 262K | ReasoningTool callStructured | - | input $0.6/M · output $3/M cache read $0.1/M | Yes | ||
| Auriko | kimi-k2.5 | openai_chat | 262K | 262K | AttachmentsReasoningTool callStructured | - | input $0.5/M · output $2.8/M | Yes | ||
| Azure | kimi-k2.5 | openai_responses | 262K | 262K | ReasoningTool callStructuredTemperature | Interleaved output | input $0.6/M · output $3/M | Yes | ||
| Neuralwatt | moonshotai/Kimi-K2.5 | openai_chat | 262K | 262K | AttachmentsReasoningTool callTemperature | - | input $0.52/M · output $2.59/M cache read $0.13/M | Yes | ||
| AIHubMix | kimi-k2.5 | openai_chat | 262K | 262K | AttachmentsReasoningTool callStructured | - | input $0.6/M · output $3/M cache read $0.1/M | Yes | ||
| TensorX | moonshotai/kimi-k2.5 | openai_chat | 262K | 262K | AttachmentsReasoningTool callStructured | - | input $0.5/M · output $2.8/M cache read $0.125/M · cache write $0.625/M | Yes | ||
| Kilo Gateway | moonshotai/kimi-k2.5 | openai_chat | 262K | 262K | AttachmentsReasoningTool callStructuredTemperature | - | input $0.6/M · output $3/M cache read $0.1/M | Yes | ||
| Tencent Coding Plan (China) | kimi-k2.5 | openai_chat | 262K | 262K | AttachmentsReasoningTool callTemperature | - | input $0/M · output $0/M cache read $0/M · cache write $0/M | Yes | ||
| Charm Hyper | kimi-k2.5 | openai_chat | 262K | 262K | Tool callStructured | - | input $0.54/M · output $2.85/M cache write $0.27/M | Yes | ||
| Vercel AI Gateway | moonshotai/kimi-k2.5 | openai_chat | 262K | 262K | AttachmentsReasoningTool callStructuredTemperature | Interleaved output | input $0.6/M · output $3/M cache read $0.1/M | Yes | ||
| Alibaba (China) | kimi-k2.5 | custom | 262K | 262K | ReasoningTool callTemperature | - | input $0.574/M · output $2.411/M | Yes | ||
| NovitaAI | moonshotai/kimi-k2.5 | openai_chat | 262K | 262K | AttachmentsReasoningTool callStructuredTemperature | - | input $0.6/M · output $3/M cache read $0.1/M | Yes | ||
| OpenRouter | moonshotai/kimi-k2.5 | openai_chat | 262K | 262K | AttachmentsReasoningTool callStructuredTemperature | Hugging Face: moonshotai/Kimi-K2.5 | input $0.57/M · output $2.85/M cache read $0.095/M | Yes | ||
| Hugging Face | moonshotai/Kimi-K2.5 | openai_chat | 262K | 262K | AttachmentsReasoningTool callTemperature | - | input $0.6/M · output $3/M cache read $0.1/M | Yes | ||
| SiliconFlow (China) | Pro/moonshotai/Kimi-K2.5 | openai_chat | 262K | 262K | ReasoningTool callStructuredTemperature | - | input $0.45/M · output $2.25/M cache read $0.07/M | Yes | ||
| Deep Infra | moonshotai/Kimi-K2.5 | openai_chat | 262K | 262K | AttachmentsReasoningTool callStructuredTemperature | - | input $0.45/M · output $2.25/M cache read $0.07/M | Yes | ||
| Cloudflare AI Gateway | workers-ai/@cf/moonshotai/kimi-k2.5 | openai_chat | 262K | 262K | AttachmentsReasoningTool callStructuredTemperature | - | input $0.6/M · output $3/M cache read $0.1/M | Yes | ||
| Alibaba Token Plan (China) | kimi-k2.5 | openai_chat | 262K | 262K | AttachmentsReasoningTool callStructuredTemperature | - | input $0/M · output $0/M cache read $0/M · cache write $0/M | Yes | ||
| Qiniu | moonshotai/kimi-k2.5 | openai_chat | 262K | 262K | AttachmentsTool callTemperature | - | input ¥0.004/K · output ¥0.021/K | Yes | ||
| Ollama Cloud | kimi-k2.5 | openai_chat | 262K | 262K | AttachmentsReasoningTool call | - | - | Yes | ||
| Jiekou.AI | moonshotai/kimi-k2.5 | openai_chat | 262K | 262K | AttachmentsReasoningTool callStructuredTemperature | - | input $0.6/M · output $3/M | Yes | ||
| Moonshot AI | kimi-k2.5 | custom | 262K | 262K | ReasoningTool callStructured | - | input $0.6/M · output $3/M cache read $0.1/M | Yes | ||
| Meganova | moonshotai/Kimi-K2.5 | openai_chat | 262K | 262K | ReasoningTool callTemperature | - | input $0.45/M · output $2.8/M | Yes | ||
| Cortecs | kimi-k2.5 | openai_chat | 262K | 262K | ReasoningTool callTemperature | - | input $0.55/M · output $2.76/M | Yes | ||
| HPC-AI | moonshotai/kimi-k2.5 | openai_chat | 262K | 262K | AttachmentsReasoningTool callStructuredTemperature | - | input $0.6/M · output $3/M cache read $0.1/M | Yes |