Introduction
Documents in. Structured data out. A small, focused API for turning unstructured text into typed JSON your code can use directly.
Untangled API takes a document — text, an uploaded file, or a URL — and returns structured data that matches a schema. No prompt engineering, no cleanup loops, no fragile regex.
What you get
- Prebuilt schemas for the most common document shapes — invoices, receipts, contracts, job postings, meeting notes.
- Custom schemas when the prebuilt set isn't enough. Pass a JSON Schema, get back data that conforms to it.
- Batch processing for high-volume pipelines. Submit up to 1000 documents at once.
- Predictable pricing. Heuristic Effort Units (HEU) — see Models — so a long contract costs more than a short receipt, and you can see exactly how much before you commit.
Where to go next
- New here? Start with the Quickstart — under five minutes from signup to first response.
- Building against a specific document type? Jump to Extract for prebuilt schemas, or Custom for your own JSON Schema.
- Looking for diff, analyze, or generate endpoints? They aren't shipped yet — see the Roadmap and join the waitlist.
Status
This is v0.1. The /v1/extract/* surface is stable and rate-limited per plan. The /v1/diff/*, /v1/analyze/*, and /v1/generate/* surfaces are on the roadmap and gated on customer demand.