Open research in AI, for all of humanity.
Forall Labs is an open research community. Anyone can join, every result is shared, and the work is aimed at AI that helps people, not just the few who can afford it.
- Open by default. Papers, code, data and models are released openly, so anyone can check, reuse and build on them.
- For all of humanity. We pick problems by who they help. Safety and real-world benefit come before hype.
- Independent and democratic. No university, company or investor decides what we study. Members propose and choose the work together.
- Credit for everyone. Every contributor is named. You do not need a title, a lab or a PhD to do serious research here.
We focus on AI itself, and apply it wherever it can do the most good.
| Theme | What it covers |
|---|---|
| Open models and tools | Models, datasets and training recipes that anyone can reproduce on modest hardware |
| AI for people | Health, education, climate and public services, built with the communities who use it |
| Safe and accountable AI | Evaluation, transparency and safety research that keeps powerful systems answerable to the public |
| Embodied and efficient AI | Robots and agents that work in the open world, and training methods small labs can afford |
First projects and papers will be listed here as they are released.
Students, independent researchers, engineers and domain experts are all welcome.
- Become a member. Membership is free and open to everyone.
- Propose a project. Pitch a research idea and find collaborators.
- Join the reading group. A weekly online session on a recent paper, open to every level.
- Support the work. Donate compute, credits or funding so serious research is not limited to the largest labs.
Questions: use the contact form.
Code is released under permissive open-source licenses. Papers, documentation and website text are CC BY 4.0. Models and datasets are released openly with their provenance documented.
∀ is the mathematical symbol for "for all".