kimi-k2.5
moonshot/kimi-k2.5
Context
262K
Max output
262K
Input / 1M
$0.60
Output / 1M
$3.00
Cached / 1M
$0.10
Cutoff
—
About this model
Kimi K2.5 is Moonshot AI's multimodal model supporting text, image, and video input, with thinking and non-thinking modes, and both dialogue and agent tasks. Context length 256K with support for long thinking and deep reasoning. Supports automatic context caching, ToolCalls, JSON Mode, Partial Mode, and internet search.
Best suited for
- Long-context workflows up to 256K tokens with deep reasoning
- Multimodal workflows (text + image + video inputs)
- Agentic task execution and dialogue
- Code generation, debugging, and refactoring
- Tool-calling agents with JSON/structured outputs
- Internet-search-augmented question answering
Built-in tools
Hosted by the gateway — enable them per request without wiring your own endpoint.
Capabilities
Vision
Accepts images alongside text in the same message.
Files
Accepts file attachments — PDFs, transcripts, spreadsheets.
Tools
Native function calling, so agents can invoke your endpoints.
System prompt
Honours a dedicated system role, separate from the user turn.
Supported parameters
max_tokensThe maximum number of tokens to generate for the chat completion. Maps to max_completion_tokens.
toolsLists tool definitions or capabilities available to the model.
tool_choiceDecides whether to use tools or just the model for generating responses.
response_typeadditional_stop_sequencesStop SequencesUp to 5 stop sequences that halt generation when fully matched. Maps to stop.
streamSends the response in real-time as it's being generated.