Mistral AI · open-weights model
Mistral-7B-Instruct-v0.3
Mistral-7B-Instruct-v0.3 is a 7.3B-parameter model by Mistral AI. Its smallest quant (Q4_K_M) needs at least 5.9GB of GPU-addressable memory at 8K context; it loads cleanly at a quality quant on 50 of the 50 machines we track.
- Parameters
- 7.3B
- Layers
- 32
- Max context
- 32K
- Attention
- GQA
- License
- apache-2.0
Config from mistralai/Mistral-7B-Instruct-v0.3
Quants and what they need
Measured file sizes; “needs” = weights + KV cache at 8K context + overhead.
| Quant | File size | Needs ≥ |
|---|---|---|
| Q8_0 | 7.7 GB | 9.3 GB |
| Q6_K | 6.0 GB | 7.5 GB |
| Q5_K_M | 5.1 GB | 6.7 GB |
| Q4_K_M | 4.4 GB | 5.9 GB |
What runs Mistral-7B-Instruct-v0.3
Every machine we track, graded at its best clean-fit quant.
Verdicts computed from published specs and measured GGUF sizes — how the math works.