---
library_name: mlx
tags:
- mlx
- oq
- quantized
---

# Qwythos-9B-v2-MLX-oQ8-mtp-draft

This model was quantized using [oQ](https://github.com/jundot/omlx) (oMLX v0.5.4.dev1) mixed-precision quantization.

## Quantization details

- **Model type**: qwen3_5_mtp
- **Bits**: 8
- **Group size**: 64
- **Format**: MLX safetensors