Everything structured data extraction needs in one tool. Parse any document with automatic image extraction, agentic exploration that reads and understands autonomously — or pick from 7 fixed strategies when you need predictable costs. Works with any LLM provider you choose. CLI for shell pipelines that compose with jq and curl, plus a TypeScript SDK for direct integration. Built-in schema validation retries until output matches your spec. Runs locally so documents never leave your machine.