Kimi VL A3B Thinking
Developed by Moonshot AI, Kimi VL A 3B Thinking is an open-source Mixture-of-Experts vision-language model that excels in advanced multimodal reasoning, long-context understanding, and agent capabilities. It is genuinely best at handling multi-turn agent interaction tasks and diverse challenging vision language tasks, including image and video comprehension, OCR, and mathematical reasoning.
Input
Output
Context
131K
Max Output
32K
Parameters
16.4B
Technical Specifications
Model TypeChat
Context Window131,072 tokens
Max Output Tokens32,000 tokens
Parameters16.4B
Release DateApr 9, 2025
Training CutoffDec 31, 2024
Licensemit
Open SourceYes
Input Modalities
TextImage
Output Modalities
Text
Capabilities
Estimated GPU Requirements for Kimi VL A3B Thinking
Estimates based on INT8 quantization. Actual requirements vary by framework and configuration.
Related Tools
Data sourced from official provider APIs and documentation
Last updated: Jun 24, 2026
Start building with the right model.
Automatically route workloads to the right model for every task, every time.