Check my GPU

LLMs ·

Kimi K2 Thinking (1T)

Parameters
1,026B (32.0B active)
Architecture
Mixture of experts
Max context
256K tokens
License
modified-mit
Released
2025-11

How to run it

Ollama

Ollama only offered this as a cloud model, and has since retired it. To run it locally, use the GGUF files below with llama.cpp or LM Studio.

Quantized files · GGUF

Works in LM Studio, llama.cpp, Jan and Ollama. No sign-in needed.

huggingface.co/unsloth/Kimi-K2-Thinking-GGUF (opens in a new tab)

ollama run hf.co/unsloth/Kimi-K2-Thinking-GGUF

Original weights · Hugging Face

Full-precision files from Moonshot AI, for fine-tuning, vLLM, Transformers.

huggingface.co/moonshotai/Kimi-K2-Thinking (opens in a new tab)

No sign-in needed.

Official repo · GitHub

Reference code, inference scripts and docs from Moonshot AI. The weights themselves download from Hugging Face.

github.com/MoonshotAI/Kimi-K2 (opens in a new tab)

git clone https://github.com/MoonshotAI/Kimi-K2

Links checked 7 days ago against Hugging Face, Ollama and GitHub.

Memory needed

Weights plus KV cache and runtime overhead. Leave a little extra free for the operating system and other apps.

KV cache: latent (MLA) KV cache.

Quantization Quality Weights Total at 8K context Total at 32K context Download
Q2_K Low 411 GB 412 GB 413 GB GGUF · folder (opens in a new tab)
Q3_K_M Low 503 GB 504 GB 506 GB GGUF · folder (opens in a new tab)
Q4_K_M Good 626 GB 627 GB 629 GB GGUF · folder (opens in a new tab)
Q5_K_M High 729 GB 730 GB 732 GB GGUF · folder (opens in a new tab)
Q6_K High 842 GB 843 GB 845 GB GGUF · folder (opens in a new tab)
Q8_0 Very high 1,088 GB 1,089 GB 1,091 GB GGUF · folder (opens in a new tab)
F16 Full 2,053 GB 2,054 GB 2,056 GB GGUF · folder (opens in a new tab) Original weights (opens in a new tab)

Best hardware for Kimi K2 Thinking (1T)

Grade · points Hardware Quant Memory Speed Fit
58 ptsGood (B) · 58 of 100 points: Model quality 29.3/45, Memory fit 4.0/25, Speed 10.3/15, Memory use 15.0/15 Apple M3 Ultra (512 GB) Q2_K 442 GB / 384 GB 18.9 tok/s Heavy offload

Check it on your own hardware

Estimates come from published hardware specs and our open formulas, not from running the model on your machine. Real speed varies with runtime, drivers and settings. Numbers marked “Measured” come from a linked benchmark.