HUD has launched DataVendor in public beta, a marketplace that lets companies sell their private data to AI labs for model training and evaluation.

It lets labs buy material not available on the public internet, like production codebases with years of Git history, after checking its quality and without seeing the raw files.

Public repositories cannot be listed or priced, since models have already trained on them.

Each listing goes through an automated scan for leaked secrets and personal data before it goes live.

For tasksets, a buyer can run one random task in its environment inside their own HUD workspace before deciding whether to buy the full set.