drip.new

We replace "rewarded ads" in free-to-play games and apps to train LLMs. Instead of showing ads, we offer short labeling tasks. This creates massive, scalable training data with real human input.