Zero-Mistral needs about 16.2 GB of memory at Q4_K_M (a 14.3 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 13 quants, from 12.9 GB (IQ4_XS) to 47.2 GB (F16). Longer context needs more: at 32K tokens the KV cache alone adds about 5.4 GB.
Zero-Mistral is a 23.6B-parameter Mistral model in GGUF format, quantized by ZeroAgency. Its trained context length is 128K tokens. The repository has 13 quant files, from 12.9 GB (IQ4_XS) to 47.2 GB (F16).
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| IQ4_XS | 12.9 GB | 14.7 GB | Too big | Get file → | |
| IQ4_NL | 13.6 GB | 15.4 GB | Too big | Get file → | |
| IQ4_XS | 13.7 GB | 15.5 GB | Too big | Get file → | |
| Q4_K_M | 14.3 GB | 16.2 GB | Too big | Get file → | |
| IQ4_NL | 14.4 GB | 16.2 GB | Too big | Get file → | |
| Q4_K_M | 15.1 GB | 17.0 GB | Too big | Get file → | |
| Q6_K | 19.3 GB | 21.2 GB | Too big | Get file → | |
| Q6_K_L | 20.1 GB | 22.0 GB | Too big | Get file → | |
| Q8_0 | 25.1 GB | 26.9 GB | Too big | Get file → | |
| Q8_0 | 25.1 GB | 26.9 GB | Too big | Get file → | |
| Q8_0 | 25.7 GB | 27.5 GB | Too big | Get file → | |
| BF16 | 47.2 GB | 49.0 GB | Too big | Get file → | |
| F16 | 47.2 GB | 49.0 GB | Too big | Get file → |
What GPU do you need for Zero-Mistral?
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: the best quant that fits is IQ4_XS (13.7 GB file, about 15.5 GB in use).
- 24 GB VRAM: the best quant that fits is Q6_K_L (20.1 GB file, about 22.0 GB in use).
- 32 GB VRAM: the best quant that fits is Q8_0 (25.7 GB file, about 27.5 GB in use).
- 48 GB VRAM: the best quant that fits is Q8_0 (25.7 GB file, about 27.5 GB in use).
Which Zero-Mistral quant should you choose?
Q4_K_M is the usual starting point for Zero-Mistral: it balances file size and quality well. Here it is a 14.3 GB file.
The Q8_0 file is 25.1 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 IQ4_XS at 12.9 GB, with the quality trade-offs that come with fewer bits. The usual sweet spot between file size and quality, and the quant most people start with.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about Zero-Mistral
How much VRAM does Zero-Mistral need?
About 16.2 GB at Q4_K_M with an 8K context window. That is the 14.3 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 Zero-Mistral on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 14.7 GB, so you would need a larger GPU or CPU offloading.
What is the file size of Zero-Mistral GGUF?
It depends on the quant: IQ4_XS is 12.9 GB, IQ4_NL is 13.6 GB, IQ4_XS is 13.7 GB, Q4_K_M is 14.3 GB, IQ4_NL is 14.4 GB, Q4_K_M is 15.1 GB and 7 more.
What context length does Zero-Mistral support?
The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 1.3 GB at 8K and 5.4 GB at 32K.
Where can I download Zero-Mistral GGUF?
The files are on Hugging Face in the ZeroAgency/Zero-Mistral-24B-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 Zero-Mistral 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.