Mimo-V2.5 needs about 208.5 GB of memory at Q4_K_M (a 190.9 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 22 quants, from 92.5 GB (IQ1_M) to 619.6 GB (BF16). Longer context needs more: at 32K tokens the KV cache alone adds about 68.5 GB.
Omnimodal 310B MiMo-V2.5 model with vision, audio, video, and long‑context support
MiMo-V2.5 is a 310B parameter Mixture of Experts model. It supports text, image, video, and audio modalities and a context length of up to 1M tokens. The model is released under the MIT license and is quantized by unsloth.
Good at
- Omnimodal perception
- Long‑context reasoning
- Agentic workflows
Limits
- No vision support in the base model
- Requires large VRAM for high‑quant versions
- License is MIT
Best for
- Multimodal chat assistants
- Long‑form content analysis
- Agentic task automation
Who it’s for. Developers building multimodal AI assistants or agents that need to process text, images, video, and audio in a single model. The model is suitable for research and production environments that can handle its memory requirements.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| IQ1_M | 92.5 GB (3 parts) | 110.1 GB | Too big | Get file → | |
| IQ2_XXS | 96.5 GB (3 parts) | 114.1 GB | Too big | Get file → | |
| IQ2_M | 96.6 GB (3 parts) | 114.2 GB | Too big | Get file → | |
| Q2_K_XL | 103.0 GB (4 parts) | 120.6 GB | Too big | Get file → | |
| IQ3_S | 114.9 GB (4 parts) | 132.5 GB | Too big | Get file → | |
| IQ3_XXS | 125.9 GB (4 parts) | 143.5 GB | Too big | Get file → | |
| Q3_K_M | 140.0 GB (4 parts) | 157.6 GB | Too big | Get file → | |
| Q3_K_XL | 140.2 GB (4 parts) | 157.8 GB | Too big | Get file → | |
| IQ4_XS | 149.4 GB (4 parts) | 167.1 GB | Too big | Get file → | |
| IQ4_NL | 152.6 GB (5 parts) | 170.2 GB | Too big | Get file → | |
| Q4_K_S | 178.8 GB (5 parts) | 196.5 GB | Too big | Get file → | |
| MXFP4_MOE | 185.4 GB (5 parts) | 203.0 GB | Too big | Get file → | |
| Q4_K_M | 190.9 GB (5 parts) | 208.5 GB | Too big | Get file → | |
| Q4_K_XL | 191.6 GB (5 parts) | 209.2 GB | Too big | Get file → | |
| Q5_K_S | 216.3 GB (6 parts) | 233.9 GB | Too big | Get file → | |
| Q5_K_M | 230.2 GB (6 parts) | 247.8 GB | Too big | Get file → | |
| Q5_K_XL | 230.7 GB (6 parts) | 248.4 GB | Too big | Get file → | |
| Q6_K | 257.0 GB (7 parts) | 274.6 GB | Too big | Get file → | |
| Q6_K_XL | 281.4 GB (7 parts) | 299.0 GB | Too big | Get file → | |
| Q8_0 | 329.3 GB (8 parts) | 346.9 GB | Too big | Get file → | |
| Q8_K_XL | 337.4 GB (8 parts) | 355.0 GB | Too big | Get file → | |
| BF16 | 619.6 GB (14 parts) | 637.3 GB | Too big | Get file → |
For 8 GB VRAM, use IQ1_M (file size 92.52 GB, estimated 110.1 GB memory at 8 k context). For 16 GB VRAM, Q4_K_S (178.84 GB, 196.5 GB memory) or Q4_K_M (190.92 GB, 208.5 GB memory) are suitable. For 24 GB VRAM, Q5_K_M (230.18 GB, 247.8 GB memory) or Q6_K (256.99 GB, 274.6 GB memory) can be used.
What GPU do you need for Mimo-V2.5?
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-V2.5 quant should you choose?
Q4_K_M is the usual starting point for Mimo-V2.5: it balances file size and quality well. Here it is a 190.9 GB file.
The Q8_0 file is 329.3 GB, about 1.7 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 IQ1_M at 92.5 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
Source
Frequently asked questions about Mimo-V2.5
How much VRAM does Mimo-V2.5 need?
About 208.5 GB at Q4_K_M with an 8K context window. That is the 190.9 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-V2.5 on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 110.1 GB, so you would need a larger GPU or CPU offloading.
What is the file size of Mimo-V2.5 GGUF?
It depends on the quant: IQ1_M is 92.5 GB, IQ2_XXS is 96.5 GB, IQ2_M is 96.6 GB, Q2_K_XL is 103.0 GB, IQ3_S is 114.9 GB, IQ3_XXS is 125.9 GB and 16 more.
What context length does Mimo-V2.5 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-V2.5 GGUF?
The files are on Hugging Face in the unsloth/MiMo-V2.5-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-V2.5 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-V2.5 on 8 GB VRAM?
Use the IQ1_M quant, which estimates 110.1 GB memory at 8 k context. This is beyond 8 GB, so it will not run on 8 GB VRAM.
What is the maximum context length of MiMo-V2.5?
The model supports up to 1M tokens of context.
Is MiMo-V2.5 licensed for commercial use?
The model is released under the MIT license, which allows commercial use.