| | --- |
| | title: README |
| | emoji: 👁 |
| | colorFrom: yellow |
| | colorTo: green |
| | sdk: static |
| | pinned: false |
| | license: apache-2.0 |
| | --- |
| | |
| | <p align="center"> |
| | <img src="https://cdn-uploads.huggingface.co/production/uploads/685ea8ff7b4139b6845ce395/1z7OO6Xv_EWEHUDemqSL1.png" alt="logo" width="200"> |
| | </p> |
| |
|
| | <p align="center"> |
| | <em>Thinking • Agentic • Research</em> |
| | </p> |
| |
|
| | --- |
| |
|
| | We are a small organization dedicated to training models for **thinking**, **agentic workflows**, and **research-oriented applications**. |
| |
|
| | Our mission is to build **small yet capable AIs** for research purposes, while exploring the true potential of compact models in advanced reasoning and autonomous tasks. |
| |
|
| | ### What we focus on |
| |
|
| | - **Thinking models** for reasoning and structured problem-solving |
| | - **Agentic models** for autonomous workflows and multi-step tasks |
| | - **Research models** for experimentation, evaluation, and exploration of small-model intelligence |
| |
|
| | ### Vision |
| |
|
| | We believe that **small models can go much further than expected** when carefully designed, trained, and evaluated. Orion exists to push these boundaries and discover what lightweight AI systems can truly achieve. |
