MiMo V2.5 Pro needs about 672.4 GB of memory at Q4_K_M (a 622.9 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 22 quants, from 1.5 GB to 640.2 GB (Q4_1). Longer context needs more: at 32K tokens the KV cache alone adds about 196.0 GB.
MiMo V2.5 Pro – 1.02T parameter Mixture‑of‑Experts model
MiMo V2.5 Pro is a 1.02T parameter Mixture‑of‑Experts language model with 42B active parameters. It supports up to 1M token context and is optimized for long‑horizon reasoning and complex tool‑use tasks.
Good at
- Long‑context reasoning
- Complex tool‑use
- Agentic task execution
Limits
- Requires large VRAM for full precision
- No vision support
- License is MIT
Best for
- Agentic software engineering
- Long‑horizon planning
- Tool‑augmented dialogue
Who it’s for. Developers building advanced agents that need to maintain coherence over very long contexts. Researchers working on Mixture‑of‑Experts architectures and long‑context benchmarks.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| — | 1.5 GB | 51.0 GB | Too big | Get file → | |
| IQ1_S | 212.0 GB (6 parts) | 261.5 GB | Too big | Get file → | |
| IQ1_M | 236.7 GB (6 parts) | 286.2 GB | Too big | Get file → | |
| IQ2_XXS | 275.2 GB (8 parts) | 324.7 GB | Too big | Get file → | |
| IQ2_XS | 306.4 GB (8 parts) | 355.9 GB | Too big | Get file → | |
| IQ2_S | 312.2 GB (8 parts) | 361.7 GB | Too big | Get file → | |
| IQ2_M | 344.7 GB (9 parts) | 394.2 GB | Too big | Get file → | |
| Q2_K | 358.3 GB (10 parts) | 407.8 GB | Too big | Get file → | |
| Q2_K_L | 359.3 GB (10 parts) | 408.7 GB | Too big | Get file → | |
| IQ3_XXS | 427.3 GB (11 parts) | 476.8 GB | Too big | Get file → | |
| Q3_K_S | 444.8 GB (12 parts) | 494.3 GB | Too big | Get file → | |
| IQ3_XS | 466.6 GB (12 parts) | 516.1 GB | Too big | Get file → | |
| Q3_K_M | 466.7 GB (12 parts) | 516.1 GB | Too big | Get file → | |
| Q3_K_L | 486.3 GB (13 parts) | 535.7 GB | Too big | Get file → | |
| Q3_K_XL | 487.1 GB (13 parts) | 536.6 GB | Too big | Get file → | |
| IQ3_M | 488.1 GB (13 parts) | 537.5 GB | Too big | Get file → | |
| IQ4_XS | 545.6 GB (14 parts) | 595.0 GB | Too big | Get file → | |
| IQ4_NL | 577.3 GB (15 parts) | 626.8 GB | Too big | Get file → | |
| Q4_0 | 578.8 GB (15 parts) | 628.3 GB | Too big | Get file → | |
| Q4_K_S | 598.4 GB (16 parts) | 647.9 GB | Too big | Get file → | |
| Q4_K_M | 622.9 GB (17 parts) | 672.4 GB | Too big | Get file → | |
| Q4_1 | 640.2 GB (17 parts) | 689.7 GB | Too big | Get file → |
For 8 GB VRAM, use Q4_0 (file size 578.77 GB, memory 628.3 GB at 8k context). For 16 GB VRAM, Q4_K_M (622.87 GB, 672.4 GB). For 32 GB VRAM, Q3_K_L (486.25 GB, 535.7 GB). Estimates are approximate.
What GPU do you need for MiMo V2.5 Pro?
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 Pro quant should you choose?
Q4_K_M is the usual starting point for MiMo V2.5 Pro: it balances file size and quality well. Here it is a 622.9 GB file.
If memory is tight, the smallest file is the first one in the table at 1.5 GB, with the quality trade-offs that come with fewer bits.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about MiMo V2.5 Pro
How much VRAM does MiMo V2.5 Pro need?
About 672.4 GB at Q4_K_M with an 8K context window. That is the 622.9 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 Pro on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 51.0 GB, so you would need a larger GPU or CPU offloading.
What is the file size of MiMo V2.5 Pro GGUF?
It depends on the quant: file is 1.5 GB, IQ1_S is 212.0 GB, IQ1_M is 236.7 GB, IQ2_XXS is 275.2 GB, IQ2_XS is 306.4 GB, IQ2_S is 312.2 GB and 16 more.
What context length does MiMo V2.5 Pro 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 49.0 GB at 8K and 196.0 GB at 32K.
Where can I download MiMo V2.5 Pro GGUF?
The files are on Hugging Face in the bartowski/MiMo-V2.5-Pro-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 Pro 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 Pro on 8 GB VRAM?
Only with the Q4_0 quantization, which still requires about 628.3 GB of VRAM at 8k context, so it is not feasible on 8 GB VRAM.
Does MiMo V2.5 Pro support vision tasks?
No, the model has no vision capabilities.