README.md

---
title: RuVI Labs
emoji: 🚀
colorFrom: gray
colorTo: gray
sdk: static
pinned: false
---

<div align="center">

<img src="logo.png" alt="RuVI Labs logo" width="96">

<br>

<img src="Ruvi.png" alt="RuVI Labs" width="100%">

# RuVI Labs

**Build good models for people.**

<sub>A research lab from India · language models, built the honest way</sub>

<br>

![India](https://img.shields.io/badge/MADE_IN-INDIA-000000?style=for-the-badge&labelColor=000000)
![Weights](https://img.shields.io/badge/WEIGHTS-OPEN-ffffff?style=for-the-badge&labelColor=000000)
![Data](https://img.shields.io/badge/DATA-OPEN-000000?style=for-the-badge&labelColor=000000)
![Scratch](https://img.shields.io/badge/BUILT-FROM_SCRATCH-ffffff?style=for-the-badge&labelColor=000000)

</div>

---

## 01 · Who we are

RuVI Labs is a small, independent research lab based in India. We design, train and release language models end to end: architecture, data, training and inference, with no borrowed foundations.

We build things that is different
<div align="center">

*Small lab, serious work.*

</div>

---

## 02 · Pragya

<div align="center">

# P R A G Y A

**A family of language models**
**trained end to end on open data**

</div>

<!-- TODO: replace the placeholders below with real sizes, context lengths and links as they ship -->

| Model | Status | Notes |
|:--|:-:|:--|
| **Pragya Brahmã** (flagship) | planned | Open weights, open data |
| **Pragya Vayu** | 🧪 Training| Built for low-resource hardware |
| **Pragya medha** | 🧪 Planned | Medium tier |



---

## 03 · Our principles

<table>
<tr>
<td width="33%" align="center">

**◼ Open weights**

Free to use, free to build on.

</td>
<td width="33%" align="center">

**◻ Open data**

Nothing hidden in the corpus.

</td>
<td width="33%" align="center">

**Capable release**
 Release whats peak 

</td>
</tr>
</table>



## 04 · What we believe

> **Good models are *different*.**

**I.** Trained to perform differently.
**II.** Built with the motive to undo what we assume is possible in AI.
**III.** Made to achieve, and made to perform.

---



<!-- TODO: edit this list to match your real roadmap -->

---




## 07 · Quick facts

| | |
|:--|:--|
| **Lab** | RuVI Labs |
| **Founder** | Arush Kumar |
| **Home** | 🇮🇳 India |
| **Family** | Pragya |
| **License** | Open weights |
| **Mode** | Building quietly. Shipping publicly. |

---


## 08 · Partners & contributors

We do not build alone. These labs have contributed time, compute, or shared
belief in the work — and we acknowledge them here.

<div align="center">

<a href="https://huggingface.co/BananaMind">
<img src="https://cdn-avatars.huggingface.co/v1/production/uploads/69ae829a8408eeb0d7dd5491/0-2aeVpWWaWlufYtgkWtK.png" alt="BananaMind AI" width="72">
</a>
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
<a href="https://huggingface.co/Refract-Labs">
<img src="https://cdn-avatars.huggingface.co/v1/production/uploads/68f4abf8f64bb4002a21a428/IwbFIrx2Y5uKZ_IW4kyUx.jpeg" alt="Refract AI Labs" width="72">
</a>

<br>

**BananaMind AI** &nbsp;·&nbsp; **Refract AI Labs**

*With gratitude.*

</div>

## 09 · FAQ

**Are the models really trained from scratch?**
Yes. Architecture, pretraining and tooling are our own. Some Models are LoRA

**Can I use the weights commercially?**
Check the license on each model card. Our goal is permissive, open release.

**Can I contribute?**
Yes. Open a discussion on any repo here, or reach out below.

<!-- TODO: confirm the license answer once licenses are final -->

---

## 10 · Follow along

Follow the organization on Hugging Face to get notified when models and datasets drop.

<!-- TODO: add links, e.g. GitHub · X · email -->

<div align="center">

<br>

<img src="logo.png" alt="RuVI Labs" width="56">

**RUVI LABS**

*small lab, serious work.*

</div>