Top AI Repos — open-source AI, indexed and scored
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
Top AI Repos tracks AI repositories on GitHub and answers two different questions about each one: is it moving right now, and would you bet a product on it.
NEO Series: Native Vision-Language Models from First Principles
| Date | Stars |
|---|---|
| 2026-07-24 | 873 |
| 2026-07-25 | 875 |
| 2026-07-28 | 878 |
| 2026-07-30 | 878 |
| 2026-08-06 | 878 |
Today
— stars today
This week
— stars this week
This month
— stars this month
Momentum
0.0
growth rate 0.00%/day
# <p align="center"> <img src="docs/NEO_logo.jpg" style="vertical-align: -10px;" :height="40px" width="40px"> NEO Series: Native Vision-Language Models </p>
<p align="center">
<img src="docs/NEO_architecture.jpg" width="80%">
</p>
- **2026/04**: [SenseNova-U1 with NEO-unify Architecture](https://arxiv.org/abs/2605.12500) (Technical Report 2026)
- **2026/02**: [NEO-unify: Building Native Multimodal Unified Models End to End](https://huggingface.co/blog/sensenova/neo-unify) (HuggingFace Blog 2026)
- **2025/12**: [From Pixels to Words -- Towards Native One-Vision Models at Scale](http://arxiv.org/abs/2605.28820) (Arxiv 2026)
- **2025/09**: [From Pixels to Words -- Towards Native Vision-Language Primitives at Scale](https://arxiv.org/abs/2510.14979) (ICLR 2026)
## 📜 News
[2026/05] 🔥 The [paper](http://arxiv.org/abs/2605.28820), [weights](https://huggingface.co/collections/Paranioar/neo1-5), and [evaluation code](https://github.com/EvolvingLMMs-Lab/NEO/blob/main/VLMEvalKit_ov/README.md) of **NEO-ov** are released !
[2025/12] 💥 **NEO-ov** has been completed !
[2026/01] 🔥 The [training code](https://github.com/EvolvingLMMs-Lab/NEO/blob/main/VLMTrainKit/README.md) of NEO is released !
[2025/10] 🔥 The [paper](https://arxiv.org/abs/2510.14979), [weights](https://huggingface.co/collections/Paranioar/neo1-0), and [evaluation code](https://github.com/EvolvingLMMs-Lab/NEO/blob/main/VLMEvalKit/README.md) of **NEO** are released !
[2025/09] 💥 **NEO** has been completed !
## 📋 Todo List
- [x] [Evaluation guide (NEO-ov)](VLMEvalKit_ov/README.md)
- [x] [Evaluation guide (NEO)](VLMEvalKit/README.md)
- [x] [Training guide](VLMTrainKit/README.md)
## 🤖 Model Zoo
We release 2B and 9B NEO-series models across different training stages.
<div style="overflow-x:auto;">
<table border="1" cellspacing="0" cellpadding="6" style="white-space:nowrap; width:fit-content; display:inline-table;">
<tr>
<th align="center">Model Name</th>
<th align="center">Model Weight</th>
</tr>
<tr>
<td>NEO</td>
<td>
<a href="https://huggingface.co/collections/Paranioar/neo1-0">
<img src="./VLMEvalKit_ov/assets/huggingface_logo.svg" width="16" height="16" /> NEO1_0 HF link
</a>
</td>
</tr>
<tr>
<td>NEO-ov</td>
<td>
<a href="https://huggingface.co/collections/Paranioar/neo1-5">
<img src="./VLMEvalKit_ov/assets/huggingface_logo.svg" width="16" height="16" /> NEO1_5 HF link
</a>
</td>
</tr>
</table>
</div>
## 💡 Motivation
<p align="center">
<img src="docs/NEO_motivation.jpg">
</p>
- **What constraints set native VLMs apart from modular ones, and to what extent can they be overcome?**
- **How to make native VLMs more accessible and democratized, thereby accelerating their progress?**
## 💡 Highlights
- 🔥 **Native Architecture:** NEO innovates a native VLM primitive that unifies pixel-word encoding, alignment, and reasoning within an encoder-free dense model architecture.
- 🔥 **Superior Efficiency:** Using under 10% of image-text data, NEO develops strong visual perception from scratch, rivaling top-tier modular VLMs and outperforming native encoder-free ones.
- 🔥 **Promising Roadmap:** NEO pioneers a promising route for scalable and powerful native encoder-free VLMs, paired with diverse reusable components that foster a cost-effective and extensible ecosystem.
## 🚀 Performance
#### 🧠 Single-Image Understanding
<p align="center">
<img src="docs/single_image_performance.png">
</p>
#### 🎬 Multi-Image & Video Understanding
<p align="center">
<img src="docs/multi_image_video_performance.png">
</p>
#### 🌐 Spatial Intelligence
<p align="center">
<img src="docs/spatial_intelligence_performance.png">
</p>
## ✒️ Citation
If **NEO series** is helpful for your research, please consider **star** ⭐ and **citation** 📝 :
```bibtex
@article{Diao2025NEO,
title = {From Pixels to Words--Towards Native Vision-Language Primitives at Scale}Excerpt of 5,480 characters
Read on GitHub
6
2
Octopus · Independent Developer · United States
1
Would you bet a product on this? Bounded 0–100 and slow moving.
matched fp:5709fda80dec5637, topic:multimodal, topic:vlm, readme:multimodal