Magika

Magika is a novel AI-powered file type detection tool developed by Google that uses a custom, highly optimized deep-learning model to accurately identify binary and textual file content types within milliseconds, even on a single CPU. Trained and evaluated on a dataset of ~100M files across 200+ content types, it achieves ~99% average precision and recall, outperforming traditional signature-based and heuristic tools—especially on textual formats like source code and configuration files. It is available as a Rust-based CLI, a Python API, and bindings for JavaScript/TypeScript and GoLang, and is used internally at Google to route files in Gmail, Drive, and Safe Browsing.