Enables APIs for LLMs to Access the Internet in Real-Time

IBL News | New York launched this month a set of APIs aimed to give Llama 2 and every other LLM real-time access to the internet, obtain up-to-date context, and augment questions from users.

Most LLMs are trained on publicly available, static data scraped from public web pages, e-books, and elsewhere. That’s sufficient to get them to perform tasks from writing emails to drafting letters and essays. However, it limits the LLMs’ knowledge to the data’s time range.

“When you ask about a recent event, like a Super Bowl score on the day of the Super Bowl, our API will search for those scores on the web, and then you can add that information, in that moment, to the LLM, and it can then use it to answer your question more accurately,” CEO and founder Richard Socher told TechCrunch.

LlamaIndex, Anthropic, and Cohere have already integrated it with their models.

Starting at $100 per month for 14,200 API calls, provides three “flavors” of API at launch: Web search, news results, and RAG (retrieval-augmented generation).