Minimax-M2.1 needs about 140.9 GB of memory at Q4_K_M (a 138.3 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 27 quants, from 55.7 GB (TQ1_0) to 457.5 GB (BF16). Longer context needs more: at 32K tokens the KV cache alone adds about 8.3 GB.
MiniMax-M2.1 – 230 billion‑parameter agentic model
MiniMax-M2.1 is a 230 billion‑parameter model built for coding, tool use, instruction following, and long‑horizon planning. It is released as a GGUF model with multiple quantizations and supports up to 196 608 context tokens.
Good at
- Robust coding and tool use
- Multilingual instruction following
- Long‑horizon planning
Limits
- Requires large VRAM for full precision
- No vision support
- License is Modified‑MIT
Best for
- Developing autonomous coding agents
- Building multilingual software workflows
- Executing complex multi‑step office tasks
Who it’s for. Engineers and researchers who need a large, open‑source model for agentic tasks. The model is suitable for local deployment with sufficient GPU memory.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| TQ1_0 | 55.7 GB | 58.3 GB | Too big | Get file → | |
| IQ1_S | 63.1 GB (2 parts) | 65.6 GB | Too big | Get file → | |
| IQ1_M | 68.4 GB (2 parts) | 71.0 GB | Too big | Get file → | |
| IQ2_XXS | 74.1 GB (2 parts) | 76.6 GB | Too big | Get file → | |
| IQ2_M | 78.1 GB (2 parts) | 80.7 GB | Too big | Get file → | |
| Q2_K | 83.3 GB (2 parts) | 85.9 GB | Too big | Get file → | |
| Q2_K_L | 83.4 GB (2 parts) | 86.0 GB | Too big | Get file → | |
| Q2_K_XL | 85.9 GB (2 parts) | 88.5 GB | Too big | Get file → | |
| IQ3_XXS | 93.3 GB (2 parts) | 95.9 GB | Too big | Get file → | |
| Q3_K_S | 98.7 GB (2 parts) | 101.3 GB | Too big | Get file → | |
| Q3_K_XL | 101.3 GB (3 parts) | 103.9 GB | Too big | Get file → | |
| Q3_K_M | 109.3 GB (3 parts) | 111.9 GB | Too big | Get file → | |
| IQ4_XS | 121.9 GB (3 parts) | 124.5 GB | Too big | Get file → | |
| IQ4_NL | 129.0 GB (3 parts) | 131.6 GB | Too big | Get file → | |
| Q4_0 | 129.5 GB (3 parts) | 132.1 GB | Too big | Get file → | |
| Q4_K_S | 130.0 GB (3 parts) | 132.6 GB | Too big | Get file → | |
| Q4_K_XL | 131.3 GB (3 parts) | 133.9 GB | Too big | Get file → | |
| Q4_K_M | 138.3 GB (3 parts) | 140.9 GB | Too big | Get file → | |
| Q4_1 | 143.2 GB (3 parts) | 145.8 GB | Too big | Get file → | |
| Q5_K_S | 157.5 GB (4 parts) | 160.1 GB | Too big | Get file → | |
| Q5_K_XL | 161.8 GB (4 parts) | 164.3 GB | Too big | Get file → | |
| Q5_K_M | 162.3 GB (4 parts) | 164.9 GB | Too big | Get file → | |
| Q6_K | 187.8 GB (4 parts) | 190.3 GB | Too big | Get file → | |
| Q6_K_XL | 194.3 GB (4 parts) | 196.9 GB | Too big | Get file → | |
| Q8_0 | 243.1 GB (5 parts) | 245.7 GB | Too big | Get file → | |
| Q8_K_XL | 261.4 GB (6 parts) | 264.0 GB | Too big | Get file → | |
| BF16 | 457.5 GB (10 parts) | 460.1 GB | Too big | Get file → |
For 8 GB VRAM, use Q4_K_S or Q4_K_XL, which estimate 132.6 GB and 133.9 GB of memory at 8 k context. For 16 GB VRAM, Q5_K_S or Q5_K_XL are appropriate, estimating 160.1 GB and 164.3 GB. Full precision requires at least 245.7 GB of VRAM.
What GPU do you need for Minimax-M2.1?
These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.
- 4 GB VRAM: no quant fits at an 8K context.
- 6 GB VRAM: no quant fits at an 8K context.
- 8 GB VRAM: no quant fits at an 8K context.
- 12 GB VRAM: no quant fits at an 8K context.
- 16 GB VRAM: no quant fits at an 8K context.
- 24 GB VRAM: no quant fits at an 8K context.
- 32 GB VRAM: no quant fits at an 8K context.
- 48 GB VRAM: no quant fits at an 8K context.
Which Minimax-M2.1 quant should you choose?
Q4_K_M is the usual starting point for Minimax-M2.1: it balances file size and quality well. Here it is a 138.3 GB file.
The Q8_0 file is 243.1 GB, about 1.8 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.
If memory is tight, the smallest file is TQ1_0 at 55.7 GB, with the quality trade-offs that come with fewer bits. Lower bit counts mean smaller files and more quality loss.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about Minimax-M2.1
How much VRAM does Minimax-M2.1 need?
About 140.9 GB at Q4_K_M with an 8K context window. That is the 138.3 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.
Can I run Minimax-M2.1 on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 58.3 GB, so you would need a larger GPU or CPU offloading.
What is the file size of Minimax-M2.1 GGUF?
It depends on the quant: TQ1_0 is 55.7 GB, IQ1_S is 63.1 GB, IQ1_M is 68.4 GB, IQ2_XXS is 74.1 GB, IQ2_M is 78.1 GB, Q2_K is 83.3 GB and 21 more.
What context length does Minimax-M2.1 support?
The GGUF header lists a trained context length of 196,608 tokens (192K). The KV cache grows with the context you use: about 2.1 GB at 8K and 8.3 GB at 32K.
Where can I download Minimax-M2.1 GGUF?
The files are on Hugging Face in the unsloth/MiniMax-M2.1-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does Minimax-M2.1 use?
Hugging Face lists the license as other. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.
Can I run MiniMax-M2.1 on 8 GB VRAM?
Only with low‑precision quantizations such as Q4_K_S or Q4_K_XL, which estimate around 133 GB of memory at 8 k context. 8 GB VRAM is insufficient for full precision.
Does MiniMax-M2.1 support vision tasks?
No, the model does not include vision capabilities.
What license applies to MiniMax-M2.1?
The model is released under a Modified‑MIT license.