Developer documentation
Documents in, structured data out
Klaaro runs an AI pipeline on every upload — OCR, classification, schema generation, and extraction — then exposes the result through a REST API, official SDKs, and an MCP server for coding agents. Automate the same workflows you use in the dashboard.
Get started in minutes
- 1. Create an API key — Sign up for the free trial (no credit card), then open Team → API keys and create a key yourself. Keys use `Authorization: Bearer sk_…` with scopes `read`, `write`, and `export`. No sales request is required.
- 2. Pick a dataset — Copy its ID from the dashboard URL or list datasets via `GET /api/v1/datasets`.
- 3. Run a recipe — Upload a document, wait for the pipeline, then read or export records.
Guides
Learn the data model, integrate with code, or connect your tools.
How it works
Teams, datasets, documents, and records — plus the pipeline from upload through OCR, classification, schema, and extraction.
Quickstart
Copy-paste recipes in TypeScript, Python, and Bash. Set env vars once, upload a file, poll for completion, and export results.
SDKs
Official TypeScript and Python clients with typed models, multipart uploads, pagination, and webhook verification.
MCP
Give Cursor, Claude Code, or Copilot access to your datasets, documents, and records over Streamable HTTP.
Integrations
Ingest via upload, URL, or webhooks. Export to CSV, Excel, or JSON — and wire automation with n8n, Zapier, or Make.
For AI agents
When to call Klaaro, how to get a free trial and API keys, how to connect, and how a document moves through the pipeline.
Reference
Base URL: https://klaaro.ai/api/v1 (or your deployment origin).
Free extraction schemas
AI-optimized schemas for popular document classes, curated by Klaaro Consulting / Customer Success. Download as JSON, Zod, Pydantic, or SQL DDL.
API reference
Interactive v1 docs with try-it requests, schemas, and code samples for every endpoint.
OpenAPI spec
Machine-readable schema for codegen, Postman, and custom clients.