We are
Project-Labs-x.

We started because the most important technology ever built was being developed in secret, by a handful of companies, with no accountability to anyone.

That felt wrong. So we decided to build in the open — models, research, tools — everything published the moment it is ready, not when it is profitable.

We are a small team. We do not have a growth function. We have engineers, researchers, and designers who believe the best way to earn trust is to show your work.

Transparency first.

Every training run, every benchmark, every failure — published. We do not cherry-pick results. The field gets stronger when everyone can see what worked and what did not.

Small on purpose.

Large teams optimise for consensus. We optimise for quality. Five people who care deeply about the work beat fifty who care about the roadmap.

Capability without hype.

We do not announce models until they are ready. We do not benchmark against cherry-picked competitors. If something is not as good as we want, we say so.

Open by default.

Weights, configs, research notes — open unless there is a specific safety reason not to be. The default is public. The exception must be justified.

Long-term thinking.

We are not optimising for an acquisition. We are building for the version of this technology that exists in ten years, not the one that exists today.