---
license: mit
base_model: WeiboAI/VibeThinker-3B
pipeline_tag: text-generation
tags: [reasoning, vibethinker, qwen2, mlx, mxfp8, osaurus]
---

![Osaurus](osaurus-x-banner.png)

# VibeThinker-3B · MXFP8

Official **OsaurusAI** MXFP8 microscaling (8.25 bpw) build of [WeiboAI/VibeThinker-3B](https://huggingface.co/WeiboAI/VibeThinker-3B) (MIT) — a 3B dense reasoning model (Qwen2). Quantized by Osaurus; runs on Apple Silicon via Osaurus / mlx_lm. Emits `<think>` reasoning blocks.

## Usage
```bash
python -m mlx_lm generate --model OsaurusAI/VibeThinker-3B-MXFP8 --prompt "What is 17 + 28?"
```
Or load in [Osaurus](https://osaurus.ai).

## Provenance
- Base: [WeiboAI/VibeThinker-3B](https://huggingface.co/WeiboAI/VibeThinker-3B) © Weibo AI — MIT
- Quantization: Osaurus · MXFP8 microscaling (8.25 bpw) · eric@osaurus.ai