---
base_model: mistralai/Mistral-Small-3.1-24B-Instruct-2503
---

[EXL3](https://github.com/turboderp-org/exllamav3) quantization of [Mistral-Small-3.1-24B-Instruct-2503](https://huggingface.co/mistralai/Mistral-Small-3.1-24B-Instruct-2503), 4 bits per weight, no vision.

For vision and other bitrates: [turboderp/Mistral-Small-3.1-24B-Instruct-2503-exl3](https://huggingface.co/turboderp/Mistral-Small-3.1-24B-Instruct-2503-exl3).