---
license: mit
title: λχ Corp.
emoji: 🤖
colorFrom: gray
thumbnail: >-
  https://cdn-uploads.huggingface.co/production/uploads/67f40ac43e188ee5436eae5f/pBPmA8e8wbJkyvUf8Crms.jpeg
sdk: gradio
---

# λχ Corp.

AI research and development organization focused on efficient language models for Portuguese-speaking markets.

---

## About

**λχ Corp.** develops lightweight, open-source AI solutions optimized for resource-constrained environments. Founded by [Marius Jabami](https://github.com/mariusjabami), we specialize in creating accessible machine learning tools for emerging markets.

### Core Focus Areas

- **Language Models** — Efficient LLMs trained natively in Portuguese
- **Mobile AI** — Inference-optimized models for low-end devices
- **Open Research** — Transparent methodology and public datasets
- **Educational Tools** — Accessible ML resources and documentation

---

## Featured Projects

| Project | Description | Status |
|---------|-------------|--------|
| [lambdAI](https://huggingface.co/lxcorp/lambda-1v-1B) | 1.1B parameter reasoning model for Portuguese | Active |
| [Synap-2b](https://huggingface.co/lxcorp/Synap-2b) | 2B conversational model | Active |
| [WNL468M](https://huggingface.co/lxcorp/WNL468M) | 468M lightweight experimental model | Research |

**Technical Stack:** PyTorch, Transformers, Quantization (8-bit), LoRA fine-tuning

---

## Mission

Making AI development accessible in Portuguese-speaking Africa through:

1. **Computational Efficiency** — Models that run on entry-level hardware
2. **Native Training** — Portuguese-first approach, not translation-based
3. **Open Source** — Full code and methodology transparency
4. **Community Focus** — Documentation and knowledge sharing

---

## Resources

- **Models:** [Hugging Face Organization](https://huggingface.co/lxcorp)
- **Code:** [GitHub](https://github.com/lxcorp-ai)
- **Contact:** lxcorp.ai@gmail.com
- **Founder:** [Marius Jabami](https://linkedin.com/in/mariusjabami)

---

**λχ Corp.** — Efficient models. Real impact.