🤔👀🎬🖥️📖 Kimi-VL-A3B-Thinking-2506: A Quick Navigation
• 82
AI & ML interests
None defined yet.
Papers
Kimi K3: Open Frontier Intelligence
PerceptionBench: Evaluating Atomic Visual Perception in Multimodal Large Language Models
Articles
Team members 55 private
Open Frontier Intelligence
Moonshot's MoE LLMs with 1 trillion parameters, exceptional on agentic intellegence
-
moonshotai/Kimi-K2-Thinking
Text Generation • 1T • Updated • 47.5k • • 1.73k -
moonshotai/Kimi-K2-Instruct-0905
Text Generation • 1T • Updated • 11.8k • • 790 -
moonshotai/Kimi-K2-Instruct
Text Generation • 1T • Updated • 311k • • 3.01k -
moonshotai/Kimi-K2-Base
Text Generation • 1T • Updated • 23.9k • 306
Moonshot's efficient MoE VLMs, exceptional on agent, long-context, and thinking
-
moonshotai/Kimi-VL-A3B-Thinking-2506
Image-Text-to-Text • 16B • Updated • 10.1k • 387 -
moonshotai/Kimi-VL-A3B-Instruct
Image-Text-to-Text • 16B • Updated • 93.8k • 285 -
moonshotai/Kimi-VL-A3B-Thinking
Image-Text-to-Text • 16B • Updated • 11.2k • 455 -
moonshotai/MoonViT-SO-400M
Image Feature Extraction • 0.4B • Updated • 4.03k • 97
Moonshot's Compute-efficient MoE LLM, first Scaling Up of Muon Optimizer
Open Frontier Intelligence
Moonshot's large visual-language model
Moonshot's MoE LLMs with 1 trillion parameters, exceptional on agentic intellegence
-
moonshotai/Kimi-K2-Thinking
Text Generation • 1T • Updated • 47.5k • • 1.73k -
moonshotai/Kimi-K2-Instruct-0905
Text Generation • 1T • Updated • 11.8k • • 790 -
moonshotai/Kimi-K2-Instruct
Text Generation • 1T • Updated • 311k • • 3.01k -
moonshotai/Kimi-K2-Base
Text Generation • 1T • Updated • 23.9k • 306
Moonshot's experimental MoE model with Kimi Delta Attention
Moonshot's efficient MoE VLMs, exceptional on agent, long-context, and thinking
-
moonshotai/Kimi-VL-A3B-Thinking-2506
Image-Text-to-Text • 16B • Updated • 10.1k • 387 -
moonshotai/Kimi-VL-A3B-Instruct
Image-Text-to-Text • 16B • Updated • 93.8k • 285 -
moonshotai/Kimi-VL-A3B-Thinking
Image-Text-to-Text • 16B • Updated • 11.2k • 455 -
moonshotai/MoonViT-SO-400M
Image Feature Extraction • 0.4B • Updated • 4.03k • 97
Kimi audio 7B models
Moonshot's Compute-efficient MoE LLM, first Scaling Up of Muon Optimizer