Google has rolled out its File Search System within the Gemini API. This represents a fully - managed Retrieval - Augmented Generation (RAG) solution, which empowers developers to "ground" the Gemini model with their proprietary data. Once developers upload files, the system takes care of the entire process automatically. It segments the content into smaller chunks, creates indexes for efficient searching, and retrieves the relevant information. As a result, the Gemini model can generate responses that are not only more accurate but also highly relevant to the context, drawing from the content of private files. According to the official article, this tool boasts several key features, including automatic management of file storage. In terms of pricing, it charges $0.15 for every million indexed tokens. Moreover, it only generates embeddings when queries are made. The tool is versatile, supporting a wide range of file formats. Additionally, it comes with example usage code to help developers get started quickly.