Moonshot AI - Mixture of experts - verified
Run Kimi K3 locally
Kimi K3 is a 2.78T mixture-of-experts (104B active) model from Moonshot AI with a 1.05M-token context window, released under Kimi K3 License (commercial use is allowed with conditions set out in the license); the smallest listed download is 1,508.7 GB, the UD-Q4_K_XL GGUF build from unsloth/Kimi-K3-GGUF.
Official weights: moonshotai/Kimi-K3 - modalities: text, image, video
- Total params
- 2.78T
- Active params
- 104B
- Context
- 1.05M
- Smallest download
- 1,508.7 GB
Fits in
Smallest download at most 85% of memoryNone of the tiers here. The smallest listed download is 1,508.7 GB, more than 85% of 256 GB.
Downloads
The official weights and every quantization listed for this model, with file sizes from Hugging Face.
| Download | Format | Size | Repo |
|---|---|---|---|
Official weights | MXFP4 | 1,560.9 GB | moonshotai/Kimi-K3 |
UD-Q4_K_XLSmallest listed | GGUF | 1,508.7 GB | unsloth/Kimi-K3-GGUF |
UD-Q8_K_XL | GGUF | 1,561.2 GB | unsloth/Kimi-K3-GGUF |
GB is decimal: 1 GB = 10^9 bytes. Add memory for the context and the operating system on top of the file size.
Run it
The official pages read for this directory publish no copyable run command for Kimi K3. Download the weights from moonshotai/Kimi-K3 and follow the model card's serving instructions.
Not sure which runtime to use? Compare local LLM runtimes by platform, API and features.
License
- License
- Kimi K3 License
other - Commercial use
- Commercial use: ConditionsCommercial use is allowed with conditions set out in the license.
- Gated download
- No. The files download directly.
Worth knowing
Ships MXFP4 weights with MXFP8 activations, always thinks, and the license requires a separate agreement for model-as-a-service providers above 20 million USD over 12 months.
Released (Hugging Face repo creation date).
FAQ
How big is the Kimi K3 download?
The official MXFP4 weights in moonshotai/Kimi-K3 are 1,560.9 GB. The smallest listed download is 1,508.7 GB, the UD-Q4_K_XL GGUF build from unsloth/Kimi-K3-GGUF. Plan for more memory than the file size: the context (KV cache) and the operating system need room too.
Can I use Kimi K3 commercially?
Kimi K3 is released under Kimi K3 License, and commercial use is allowed with conditions set out in the license. The license text is at https://huggingface.co/moonshotai/Kimi-K3/blob/main/LICENSE.
How much memory do I need for Kimi K3?
Its smallest listed download is 1,508.7 GB, more than 85% of 256 GB, so it needs a multi-GPU server or a machine with more memory than the largest tier here.
What is the context length of Kimi K3?
1,048,576 tokens. A longer context needs more memory for the KV cache, so you can run with a shorter one when memory is tight.
Sources
Last verified from the pages below.
- Hugging Face model card - huggingface.co
- Hugging Face API - huggingface.co
- unsloth GGUF - huggingface.co
Newsletter
New open weights, every week
One email per week. New model releases, open-weight drops and what they mean for what you build.
Keep exploring
Best local AI models
Our curated picks by hardware tier, with benchmarks and the reasoning.
Model API pricing and specs
First-party prices per 1M tokens when you would rather call an API.
AI benchmark leaderboards
Sourced scores, for when you need to compare quality.
Setup guides
Point your coding tools at the model you picked.