MI

MiniMax M2.7 Abliterated

MinimaxGGUF by YoussofalUpdated 2026-04-14
Download on Hugging Face
Parameters229B
Context192K
Layers62
Architectureminimax-m2
Smallest quant83.3 GB
Quick answer

MiniMax M2.7 Abliterated 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 6 quants, from 83.3 GB (Q2_K) to 457.5 GB (BF16). Longer context needs more: at 32K tokens the KV cache alone adds about 8.3 GB.

229B sparse MoE model with 200K context and refusal removal

MiniMax M2.7 Abliterated is a 229B sparse MoE model with 10B active parameters per token. It offers a 200K context window and retains long‑context instruction following while removing base refusal behavior. The model is available in several GGUF quantizations for llama.cpp deployment.

Good at

  • Long‑context reasoning
  • Agent‑team collaboration
  • Tool‑based problem solving

Limits

  • No vision support
  • Requires large VRAM for higher quants
  • Non‑commercial license

Best for

  • Complex software engineering tasks
  • Large‑scale data analysis
  • Multi‑agent workflow orchestration

Who it’s for. Developers and researchers needing high‑capacity reasoning and agent collaboration in a non‑commercial setting. The model is suitable for environments that can accommodate its memory demands and comply with the non‑commercial license.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q2_K83.3 GB (2 parts) 85.9 GB
Too big Get file →
Q3_K_M109.3 GB (3 parts) 111.9 GB
Too big Get file →
Q4_K_M138.3 GB (4 parts) 140.9 GB
Too big Get file →
Q6_K187.8 GB (5 parts) 190.3 GB
Too big Get file →
Q8_0243.1 GB (5 parts) 245.7 GB
Too big Get file →
BF16457.5 GB (10 parts) 460.1 GB
Too big Get file →

For 8 GB VRAM, use Q2_K (file size 83.3 GB, estimated 85.9 GB memory). For 16 GB VRAM, Q3_K_M (109.27 GB, 111.9 GB memory) is recommended. For 32 GB VRAM, Q4_K_M (138.34 GB, 140.9 GB memory) provides a balance between quality and memory usage. BF16 and higher quants require 40 GB+ VRAM.

What GPU do you need for MiniMax M2.7 Abliterated?

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 M2.7 Abliterated quant should you choose?

Q4_K_M is the usual starting point for MiniMax M2.7 Abliterated: 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 Q2_K at 83.3 GB, with the quality trade-offs that come with fewer bits. Use it only when memory is very tight. Expect a clear drop in quality, which is usually worse on small models and easier to live with on large ones.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architectureminimax-m2
Parameters229B (approximate, mixture of experts)
Layers62
Trained context196,608 tokens
KV cache (fp16)248 KB per token

Source

Quantized byYoussofal
FamilyMinimax
Licenseother
Downloads (HF)2,941
Last updated2026-04-14

Frequently asked questions about MiniMax M2.7 Abliterated

How much VRAM does MiniMax M2.7 Abliterated 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 M2.7 Abliterated on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 85.9 GB, so you would need a larger GPU or CPU offloading.

What is the file size of MiniMax M2.7 Abliterated GGUF?

It depends on the quant: Q2_K is 83.3 GB, Q3_K_M is 109.3 GB, Q4_K_M is 138.3 GB, Q6_K is 187.8 GB, Q8_0 is 243.1 GB, BF16 is 457.5 GB.

What context length does MiniMax M2.7 Abliterated 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 M2.7 Abliterated GGUF?

The files are on Hugging Face in the Youssofal/MiniMax-M2.7-Abliterated-Heretic-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 M2.7 Abliterated 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 this model on 8 GB VRAM?

Yes, the Q2_K quant requires about 85.9 GB memory at 8k context, which is not feasible on 8 GB VRAM. A smaller quant or lower context is needed.

Does this model support vision tasks?

No, the model does not include vision capabilities.

Is commercial use allowed?

No, the license is non‑commercial; commercial use requires written authorization from MiniMax.

Other GGUFs of MiniMax-M2.7

Similar models

Browse all →

More from Youssofal

All by Youssofal →

More Minimax models

All Minimax →

Other large models

All large models →