Xberg

Xberg is a document-intelligence engine, built with a Rust core, that extracts clean text, tables, metadata and structured data from more than 100 document formats across over 140 file extensions, plus code intelligence for hundreds of programming languages. It is the next iteration of the earlier project Kreuzberg, rebuilt and rebranded as a new v1 line.

Pointed at a PDF, a scanned image, a spreadsheet, an audio file, a URL, an archive, or a source tree, it handles format detection, reading, OCR and extraction in one engine instead of a hand-assembled pipeline. It can fetch and extract a single URL or crawl and follow links, transcribe speech from audio and video through Whisper ONNX, recursively extract nested archives with zip-bomb and depth guards, run OCR through Tesseract, PaddleOCR, Candle or vision-language-model backends with fallback chains and confidence scores, and reconstruct reading order and table structure with machine-learning layout and table models.

It is callable from Rust, Python, Node.js, Go, Java, C#, Ruby, PHP, Elixir, Dart, Swift, Zig, WASM, Kotlin or C FFI, and can run as a library, a CLI tool, a REST API, or an MCP server.