MiMo V2.6 Flash RL UNCENSORED needs about 186.0 GB of memory at MXFP4 (a 168.4 GB file) with an 8K context window, which is more than a 48 GB GPU holds. Longer context needs more: at 32K tokens the KV cache alone adds about 68.5 GB.
MiMo V2.6 Flash RL UNCENSORED – 310B parameters, 1M context, uncensored MiMo model
MiMo V2.6 Flash RL UNCENSORED is a 310B‑parameter MiMo model with 1,048,576‑token context. It removes refusal behavior at the weight level while preserving coding, math, bilingual coverage, vision, audio, and video understanding. The bundle is FP8 attention + MXFP4 experts and is compatible with transformers, vLLM, and SGLang.
Good at
- High‑throughput inference with DFlash head
- Large‑context reasoning up to 1M tokens
- Uncensored instruction following
Limits
- Requires ~186GB VRAM at 8k context
- Large file size 168.42GB
- No runtime hooks or steering vectors
Best for
- High‑throughput inference
- Large‑context tasks
- Uncensored content generation
Who it’s for. Developers and researchers who need a large MiMo model without refusal behavior and with full vision, audio, and video capabilities. The MIT license allows free use and modification.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| MXFP4 | 168.4 GB (4 parts) | 186.0 GB | Too big | Get file → |
The only available quant is MXFP4, 168.42GB. It requires approximately 186GB VRAM at 8k context. For smaller memory budgets, a different quant is not provided, so this model is best suited for systems with very large GPU memory.
What GPU do you need for MiMo V2.6 Flash RL UNCENSORED?
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 MiMo V2.6 Flash RL UNCENSORED quant should you choose?
MXFP4 (168.4 GB) is the middle option in this repository. 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 MiMo V2.6 Flash RL UNCENSORED
How much VRAM does MiMo V2.6 Flash RL UNCENSORED need?
About 186.0 GB at MXFP4 with an 8K context window. That is the 168.4 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 MiMo V2.6 Flash RL UNCENSORED on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 186.0 GB, so you would need a larger GPU or CPU offloading.
What is the file size of MiMo V2.6 Flash RL UNCENSORED GGUF?
It depends on the quant: MXFP4 is 168.4 GB.
What context length does MiMo V2.6 Flash RL UNCENSORED support?
The GGUF header lists a trained context length of 1,048,576 tokens (1024K). The KV cache grows with the context you use: about 17.1 GB at 8K and 68.5 GB at 32K.
Where can I download MiMo V2.6 Flash RL UNCENSORED GGUF?
The files are on Hugging Face in the AliceThirty/MiMo-V2.6-Flash-RL-UNCENSORED-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 MiMo V2.6 Flash RL UNCENSORED use?
Hugging Face lists the license as mit. 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 MiMo V2.6 Flash RL UNCENSORED on 8 GB VRAM?
No. The MXFP4 quant requires about 186GB VRAM at 8k context, far exceeding 8 GB.
Does this model support vision and audio?
Yes. It includes a vision‑language encoder, an audio encoder, and video understanding.
Is the model compliant with safety guidelines?
Refusal behavior is removed at the weight level; the model follows instructions across task categories while maintaining safety compliance in most categories.