unsloth/MiniMax-M3-GGUF

Read our How to Run MiniMax M3 Guide!

<div> <p style="margin: 0 0 0px 0; margin-top: 0px;"> <em>See <a href="https://unsloth.ai/docs/basics/unsloth-dynamic-v2.0-gguf">Unsloth Dynamic 2.0 GGUFs</a> for our quantization benchmarks.</em> </p> <div style="display: flex; gap: 5px; align-items: center; margin-bottom: 0px;"> <a href="https://github.com/unslothai/unsloth/"> <img src="https://github.com/unslothai/unsloth/raw/main/images/unsloth%20new%20logo.png" width="133"> </a> <a href="https://discord.gg/unsloth"> <img src="https://github.com/unslothai/unsloth/raw/main/images/Discord%20button.png" width="173"> </a> <a href="https://unsloth.ai/docs/models/minimax-m3"> <img src="https://raw.githubusercontent.com/unslothai/unsloth/refs/heads/main/images/documentation%20green%20button.png" width="143"> </a> </div>

<ul style="margin: 0;"> <li>EXPERIMENTAL GGUF / support for MiniMax-M3</li> <li><b>Jun 12 Update:</b> You can now run MiniMax M3 in Unsloth Studio. See our <a href="https://unsloth.ai/docs/models/minimax-m3#unsloth-studio-guide">Guide</a>.</li> <li>Example of MiniMax M3 (5-bit GGUF) running in Unsloth Studio:</li> </ul> </div> <img width="600" alt="minimax m3 in unsloth studio" src="https://cdn-uploads.huggingface.co/production/uploads/62ecdc18b72a69615d6bd857/4ZRsRdcz9YsSkNDisoxfO.gif" />

<div style="margin: 0;"> <b>EXPERIMENTAL GGUF / support for MiniMax-M3 in llama.cpp:</b> </div> </div>

MiniMax-M3 support in llama.cpp is preliminary and not yet in a released build. To run these GGUFs, build llama.cpp from PR #24523:

git clone https://github.com/ggml-org/llama.cpp
cd llama.cpp
git fetch origin pull/24523/head:minimax-m3
git checkout minimax-m3
cmake -B build -DGGML_CUDA=ON
cmake --build build --config Release -j --target llama-cli llama-server

Then run a quant. The model is large (~428B params), so offload across GPUs with -ngl 99 or keep the weights in CPU RAM:

./build/bin/llama-cli -hf unsloth/MiniMax-M3-GGUF:UD-IQ1_M

Note: MiniMax Sparse Attention is not supported yet, so inference falls back to dense attention.

MiniMax-M3

Highlights:

• —Native Multimodality: M3 undergoes mixed-modality training from the very first step, enabling deeper semantic fusion across text, image, and video.
• —Context Scaling via Sparse Attention: M3 introduces MiniMax Sparse Attention (MSA) to improve long context efficiency. M3 delivers 9× prefill and 15× decode speedups compared to M2 at 1M context, reducing per-token compute to 1/20.
• —Coding & Cowork Capability: M3 achieves frontier-level performance across long-horizon agentic benchmarks, excelling in both coding and cowork.

Model Details

How to Use

• —MiniMax Agent
• —MiniMax API

M3 supports two reasoning modes:

• —thinking — for complex reasoning, agentic tasks, and long-horizon collaboration.
• —non-thinking — for latency-sensitive scenarios such as chat and code completion.

Local Deployment

Download the model:

hf download MiniMaxAI/MiniMax-M3 --local-dir MiniMax-M3

You can also get model weights from ModelScope.

Inference Parameters

We recommend the following parameters for best performance: temperature=1.0, top_p=0.95, top_k=40. Default system prompt:

You are a helpful assistant. Your name is MiniMax-M3 and was built by MiniMax.