Skip to main content
Doclo is a TypeScript SDK and cloud platform for building document AI pipelines. Extract structured data from PDFs, images, and other documents using multiple AI providers with built-in consensus voting, citations, and observability.

Choose Your Path

SDK Quickstart

Build document processing pipelines in your own infrastructure using the Doclo SDK.

Cloud Quickstart

Execute flows via the Doclo Cloud API using our hosted infrastructure.

REST API

Call the Cloud API directly without the SDK for integrations like n8n or Zapier.

Core Features

Multi-Provider Orchestration

Use OpenAI, Anthropic, Google, or xAI for extraction. Switch providers without changing your code.

Consensus Voting

Run multiple providers in parallel and use voting to improve accuracy on critical fields.

Citations & Traceability

Track exactly where each extracted value came from in the source document.

Observability

Built-in metrics, cost tracking, and distributed tracing for production workloads.

Packages

The SDK is modular—install only what you need:
PackageDescription
@doclo/coreCore types, utilities, and validation
@doclo/nodesProcessing nodes (parse, extract, split, etc.)
@doclo/flowsFlow builder and orchestration
@doclo/providers-llmLLM integrations (OpenAI, Anthropic, Google, xAI)
@doclo/providers-datalabOCR integrations (Surya, Marker)
@doclo/schemasPre-built schemas for common documents
@doclo/clientCloud API client

What’s Next?

Quickstart

Extract data from your first document in under 10 minutes.

Concepts

Understand how Doclo’s architecture works.