File size: 749 Bytes
83b1d62 5f331a3 e548842 dd52ff5 09f25dc dd52ff5 09f25dc |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 |
---
title: README
emoji: 🔥
colorFrom: pink
colorTo: indigo
sdk: static
pinned: false
---
**Moxin LM: From SOTA Research to Efficient Deployment**
- **Open Creation:** The **Moxin-7B series** is our truly open, SOTA-performing LLM and VLM. We build, fine-tune, and openly release our own models.
- **Efficient Deployment:** We specialize in extreme quantization, creating resource-efficient variants of popular models (like DeepSeek and Kimi) to run anywhere.
We unleash the power of reproducible AI 🚀. Explore our models below and on [GitHub](https://github.com/moxin-org), and read our research on [Moxin 7B (Open Creation)](https://arxiv.org/abs/2412.06845) and [MoE Compression (Efficient Deployment)](https://arxiv.org/abs/2509.25689). |