Documentation
Everything you need to structure enterprise documents into schema-validated data — from platform concepts to API endpoint reference.
Start here
Quickstart — ship in 5 minutes
Get a key, extract your first document, receive structured data with confidence scores. Three modes, one endpoint. Working curl included.
Platform Guide
Learn the platform
Sources, Field Registry, schemas and Specs, pipelines, cases, validation, review, and delivery, every group in one place.
Apps
Turn data into decisions
Create apps by describing them, read the verdict matrix, audit runs in the Run Inspector, and work the Human Review inbox.
API Docs
Build with the API
REST API reference: extract, run, ask, documents, schemas, Specs, pipelines, data products, apps, webhooks, and more.
Node SDK
Use TypeScript
Install @talonic/node, extract your first document, and handle retries, errors, and the CLI from Node.js.
MCP Server
Connect an AI agent
Hosted MCP server for Claude, Cursor, and any MCP client, with the full tool reference for extraction, Specs, Ask, and Apps.
Concepts
Understand the model
Field Registry, schemas, cases, provenance, the pipeline and Spec rail, and the 529-type document ontology.