LLM EXPLORER
 LLM Explorer: A Curated Large Language Model Directory and Analytics
Dark Theme
  Agents   Benchmarks   LLM  List    LLM Hosting    LLM Leaderboards    Blog    Newsfeed    Advertise

Kimi K2 Thinking 4bit by mlx-community

 »  All LLMs  »  mlx-community  »  Kimi K2 Thinking 4bit   URL Share it on

  810   

Kimi K2 Thinking 4bit is an open-source language model by mlx-community. Features: 1026.4b LLM, VRAM: 151.2GB, Context: 256K, License: other, Quantized, LLM Explorer Score: 0.25.

  4-bit   4bit Base model:moonshotai/kimi-k2-... Base model:quantized:moonshota...   Conversational   Custom code   Kimi k2   Mlx   Quantized   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/mlx-community/Kimi-K2-Thinking-4bit 

Kimi K2 Thinking 4bit Benchmarks

LLME Score: 0.24606
nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").
Kimi K2 Thinking 4bit (mlx-community/Kimi-K2-Thinking-4bit)
🌟 Advertise your project 🚀

Kimi K2 Thinking 4bit Parameters and Internals

LLM NameKimi K2 Thinking 4bit
Repository 🤗https://huggingface.co/mlx-community/Kimi-K2-Thinking-4bit 
Base Model(s)  Kimi K2 Thinking   moonshotai/Kimi-K2-Thinking
Model Size1026.4b
Required VRAM151.2 GB
Updated2026-05-05
Maintainermlx-community
Model Typekimi_k2
Model Files  4.2 GB: 1-of-180   3.5 GB: 1-of-182   3.2 GB: 2-of-180   3.5 GB: 2-of-182   3.3 GB: 3-of-180   3.5 GB: 3-of-182   3.2 GB: 4-of-180   3.8 GB: 4-of-182   3.2 GB: 5-of-180   3.5 GB: 5-of-182   3.3 GB: 6-of-180   3.5 GB: 6-of-182   3.2 GB: 7-of-180   3.8 GB: 7-of-182   3.2 GB: 8-of-180   3.5 GB: 8-of-182   3.3 GB: 9-of-180   3.5 GB: 9-of-182   3.2 GB: 10-of-180   3.8 GB: 10-of-182   3.2 GB: 11-of-180   3.5 GB: 11-of-182   3.3 GB: 12-of-180   3.5 GB: 12-of-182   3.2 GB: 13-of-180   3.8 GB: 13-of-182   3.2 GB: 14-of-180   3.5 GB: 14-of-182   3.3 GB: 15-of-180   3.5 GB: 15-of-182   3.2 GB: 16-of-180   3.8 GB: 16-of-182   3.2 GB: 17-of-180   3.5 GB: 17-of-182   3.3 GB: 18-of-180   3.5 GB: 18-of-182   3.2 GB: 19-of-180   3.8 GB: 19-of-182   3.2 GB: 20-of-180   3.5 GB: 20-of-182   3.3 GB: 21-of-180   3.5 GB: 21-of-182   3.2 GB: 22-of-180   3.8 GB: 22-of-182
Quantization Type4bit
Model ArchitectureDeepseekV3ForCausalLM
Licenseother
Context Length262144
Model Max Length262144
Transformers Version4.51.3
Vocabulary Size163840
Torch Data Typebfloat16

Best Alternatives to Kimi K2 Thinking 4bit

Best Alternatives
Context / RAM
Downloads
Likes
Kimi K2 Instruct 4bit128K / 143.2 GB465413
Kimi K2 Thinking256K / 158.2 GB505917
...i K2 Instruct 0905 Mlx DQ3 K M256K / 204.6 GB172612
Note: green Score (e.g. "73.2") means that the model is better than mlx-community/Kimi-K2-Thinking-4bit.

Rank the Kimi K2 Thinking 4bit Capabilities

🆘 Have you tried this model? Rate its performance. This feedback would greatly assist ML community in identifying the most suitable model for their needs. Your contribution really does make a difference! 🌟

Instruction Following and Task Automation  
Factuality and Completeness of Knowledge  
Censorship and Alignment  
Data Analysis and Insight Generation  
Text Generation  
Text Summarization and Feature Extraction  
Code Generation  
Multi-Language Support and Translation  

What open-source LLMs or SLMs are you in search of? 53727 in total.

Email us: info@extractum.io. Our Privacy Policy  | Terms and Conditions  | Suggest an improvement.
Our Social Media →  
Original data from HuggingFace, OpenCompass and various public git repos.
Check out Ag3ntum — our secure, self-hosted AI agent for server management.
Release v20260328a
  Support LLM Explorer