// about
About Axiomi
Axiomi is an independent directory of foundational AI: open-weight models, machine-learning frameworks, training and fine-tuning tools, interpretability research, datasets, benchmarks and research platforms.
We built it because the most important AI tools for researchers are scattered across GitHub, Hugging Face, arXiv and lab websites. Axiomi brings them together with short, original descriptions, accurate links to official sources, and public signals such as GitHub stars, fetched automatically and dated.
The research feed combines the arXiv API, Hugging Face Daily Papers and trending models, GitHub and Hacker News. It is rebuilt automatically every day, and your browser fetches the newest items live from sources that allow it.
Listings are editorial and cannot be bought. Anyone can suggest a model, library or dataset using the submission form; we review each suggestion before publishing.
How we curate
- Real tools only. Every listing is a real product or project with a working official website. Links are checked periodically and dead entries are removed.
- Original descriptions. We write each summary ourselves from official public information. We never copy other directories.
- No invented numbers. Stars, dates and discussion counts come from public APIs (GitHub, Hacker News Algolia, Hugging Face) and show when they were fetched. If a fetch fails we show nothing rather than guess.
- Independent. Placement cannot be bought. Ads, where shown, are clearly labelled and kept separate from editorial content.
What we cover
Axiomi currently indexes 56 tools across 6 categories: Datasets, Interpretability, ML Frameworks, Open-Weight Models, Research Platforms, Training & Fine-tuning.
Data sources
Live feed and statistics: Hacker News, arXiv, HF Daily Papers, HF Trending Models, GitHub. Tool logos are loaded from Google's public favicon service and belong to their respective owners.
Contact
Questions, corrections or partnership ideas? Use the contact form. Want to be listed? Submit a tool.