MiniMax - Mixture of experts - verified
Run MiniMax-M3 locally
MiniMax-M3 is a 427B mixture-of-experts (23B active) model from MiniMax with a 1.05M-token context window, released under MiniMax Community License (commercial use is allowed with conditions set out in the license); the smallest listed download is 241.5 GB, the 4bit MLX build from mlx-community/MiniMax-M3-4bit.
Official weights: MiniMaxAI/MiniMax-M3 - modalities: text, image, video
- Total params
- 427B
- Active params
- 23B
- Context
- 1.05M
- Smallest download
- 241.5 GB
Fits in
Smallest download at most 85% of memoryNone of the tiers here. The smallest listed download is 241.5 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 | BF16 | 854.2 GB | MiniMaxAI/MiniMax-M3 |
UD-Q4_K_M | GGUF | 264.9 GB | unsloth/MiniMax-M3-GGUF |
Q8_0 | GGUF | 453.6 GB | unsloth/MiniMax-M3-GGUF |
4bitSmallest listed | MLX | 241.5 GB | mlx-community/MiniMax-M3-4bit |
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 MiniMax-M3. Download the weights from MiniMaxAI/MiniMax-M3 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
- MiniMax Community 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
The card offers three reasoning modes through a thinking parameter and recommends temperature 1.0 with top_p 0.95, and the Ollama entry is cloud only.
Released (Hugging Face repo creation date).
FAQ
How big is the MiniMax-M3 download?
The official BF16 weights in MiniMaxAI/MiniMax-M3 are 854.2 GB. The smallest listed download is 241.5 GB, the 4bit MLX build from mlx-community/MiniMax-M3-4bit. Plan for more memory than the file size: the context (KV cache) and the operating system need room too.
Can I use MiniMax-M3 commercially?
MiniMax-M3 is released under MiniMax Community License, and commercial use is allowed with conditions set out in the license. The license text is at https://huggingface.co/MiniMaxAI/MiniMax-M3/blob/main/LICENSE.
How much memory do I need for MiniMax-M3?
Its smallest listed download is 241.5 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 MiniMax-M3?
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
- HF API - huggingface.co
- config.json - 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.