$42,000 for 50 million tweets. $12,000 for 50 million Reddit posts. Is this the end of public APIs?
Twitter’s public API was launched only a few months after the product was released in 2006 to combat third-party developers scraping and offering unofficial APIs. The API enabled reasonably broad access to Twitter for any developer. As a result, third-party applications flourished (like TweetDeck, Tweetbot, and Twitterriffic).
But over time, the API became more closed off. They limited the number of connected users for third-party applications that looked like Twitter. Rate limits became harsher. Eventually, they killed off third-party applications altogether.
Later, they reintroduced the API but with a new price tag. Other companies like Reddit are following suit.
Why?
Some users of these APIs collected vast amounts of data from sites like Twitter and Reddit through their APIs. This data was ultimately used to train modern large language models.
In an increasingly competitive market, third-party applications competed with the platforms themselves.
Is this the end of public APIs? Not necessarily. While Reddit and Twitter wall off their APIs, other companies will embrace openness. Some will emerge with different business models or data models made better by public APIs.
The End of Public APIs
Mainly to profit from ai models pre-training on their data