AI Models Directory
Nvidia
Open Source
Kimi K2.5
A fast decision page for teams comparing performance, cost, context window, and critical capabilities without digging through raw specs.
Max Context (In)
262K
Max Output (Out)
262K
Input (1M tokens)
?
Output (1M tokens)
?
Quick signals
- Provider: Nvidia
- Inputs: text, image, video
- Latest update: 2026-01-27
Usage profile
- Context window: 262K
- Max output: 262K
- Open weights: Yes
Capabilities
Vision
Tool Calling
Structured Output (JSON)
File Attachments
Reasoning
Open Source