mistral needs about 6.5 GB of memory at Q4_1 (a 5.0 GB file) with an 8K context window, so it fits on a 8 GB GPU. This repository has 15 quants, from 3.1 GB (Q2_K) to 14.7 GB (F16). Longer context needs more: at 32K tokens the KV cache alone adds about 4.3 GB.
mistral is a 7.24B-parameter Mistral model in GGUF format, quantized by itsdotscience. Its trained context length is 32K tokens. The repository has 15 quant files, from 3.1 GB (Q2_K) to 14.7 GB (F16).
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q2_K | 3.1 GB | 4.7 GB | Fits | Get file → | |
| Q3_K_S | 3.6 GB | 5.2 GB | Fits | Get file → | |
| Q3_K | 3.9 GB | 5.5 GB | Fits | Get file → | |
| Q3_K_L | 4.2 GB | 5.8 GB | Fits | Get file → | |
| Q4_0 | 4.5 GB | 6.1 GB | Fits | Get file → | |
| Q4_K_S | 4.6 GB | 6.1 GB | Fits | Get file → | |
| Q4_K | 4.8 GB | 6.4 GB | Fits | Get file → | |
| Q4_1 | 5.0 GB | 6.5 GB | Fits | Get file → | |
| Q5_K_S | 5.4 GB | 7.0 GB | Fits | Get file → | |
| Q5_0 | 5.4 GB | 7.0 GB | Fits | Get file → | |
| Q5_K | 5.5 GB | 7.1 GB | Fits | Get file → | |
| Q5_1 | 5.9 GB | 7.4 GB | Tight | Get file → | |
| Q6_K | 6.4 GB | 7.9 GB | Tight | Get file → | |
| Q8_0 | 8.1 GB | 9.7 GB | Too big | Get file → | |
| F16 | 14.7 GB | 16.3 GB | Too big | Get file → |
What GPU do you need for 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: the best quant that fits is Q3_K_L (4.2 GB file, about 5.8 GB in use).
- 8 GB VRAM: the best quant that fits is Q6_K (6.4 GB file, about 7.9 GB in use).
- 12 GB VRAM: the best quant that fits is Q8_0 (8.1 GB file, about 9.7 GB in use).
- 16 GB VRAM: the best quant that fits is Q8_0 (8.1 GB file, about 9.7 GB in use).
- 24 GB VRAM: the best quant that fits is F16 (14.7 GB file, about 16.3 GB in use).
Which mistral quant should you choose?
Q4_1 (5.0 GB) is the middle option in this repository. The usual sweet spot between file size and quality, and the quant most people start with.
If memory is tight, the smallest file is Q2_K at 3.1 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 mistral
How much VRAM does mistral need?
About 6.5 GB at Q4_1 with an 8K context window. That is the 5.0 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 mistral on an 8 GB GPU?
Yes, with Q6_K: the 6.4 GB file needs about 7.9 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of mistral GGUF?
It depends on the quant: Q2_K is 3.1 GB, Q3_K_S is 3.6 GB, Q3_K is 3.9 GB, Q3_K_L is 4.2 GB, Q4_0 is 4.5 GB, Q4_K_S is 4.6 GB and 9 more.
What context length does mistral support?
The GGUF header lists a trained context length of 32,768 tokens (32K). The KV cache grows with the context you use: about 1.1 GB at 8K and 4.3 GB at 32K.
Where can I download mistral GGUF?
The files are on Hugging Face in the itsdotscience/mistral-7b-v0.2-gguf repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.