Kimi K2 Base
Overview
Overview
Kimi K2 base model is a state-of-the-art mixture-of-experts (MoE) language model with 32 billion activated parameters and 1 trillion total parameters. Trained on 15.5 trillion tokens with the MuonClip optimizer, this is the foundation model before instruction tuning. It demonstrates strong performance on knowledge, reasoning, and coding benchmarks while being optimized for agentic capabilities.
Kimi K2 Base was released on July 11, 2025.
Performance
Timeline
ReleasedUnknown
Knowledge CutoffUnknown
Specifications
Parameters
1000.0B
License
MIT
Training Data
Unknown
Tags
tuning:base
Benchmarks
Benchmarks
Kimi K2 Base Performance Across Datasets
Scores sourced from the model's scorecard, paper, or official blog posts
llm-stats.com - Wed Jan 21 2026Notice missing or incorrect data?Start an Issue discussion→
Pricing
Pricing
Pricing, performance, and capabilities for Kimi K2 Base across different providers:
No pricing information available for this model.
API Access
API Access Coming Soon
API access for Kimi K2 Base will be available soon through our gateway.
Recent Posts
Recent Reviews
FAQ
Common questions about Kimi K2 Base
Kimi K2 Base was released on July 11, 2025 by MoonshotAI.
Kimi K2 Base was created by MoonshotAI.
Kimi K2 Base has 1000.0 billion parameters.
Kimi K2 Base is released under the MIT license. This is an open-source/open-weight license.