Kimi K2.5
Moonshot AI's open-source visual agentic model with native multimodal capabilities and multi-agent orchestration
Kimi K2.5
Moonshot AI • January 2026
Training Data
~15 trillion visual and text tokens
Kimi K2.5
January 2026
Parameters
1 trillion (32B active)
Training Method
Mixture of Experts with Visual Agentic Training
Context Window
256,000 tokens
Knowledge Cutoff
January 2026
Key Features
Visual Agentic • Native Multimodal • Multi-Agent Swarm • Open Source (MIT)
Capabilities
Visual Understanding: Outstanding
Agentic Tasks: Outstanding
Coding: Outstanding
What's New in This Version
Native vision-language integration with MoonViT encoder, thinking/instant modes, and self-directed multi-agent orchestration
Moonshot AI's open-source visual agentic model with native multimodal capabilities and multi-agent orchestration
What's New in This Version
Native vision-language integration with MoonViT encoder, thinking/instant modes, and self-directed multi-agent orchestration
Technical Specifications
Key Features
Capabilities
Other Moonshot AI Models
Explore more models from Moonshot AI
Kimi K2
Moonshot AI's flagship model - #1 open-source LLM on LMSYS Arena with exceptional agentic capabilities
Kimi K1.5
Moonshot AI's advanced reasoning model matching OpenAI o1 performance
Kimi Chat
Moonshot AI's flagship chatbot with industry-leading long context capabilities
