Docspeed Customer API

Build document workflows that return structured JSON and page-grounded evidence from PDFs and images.

What is Docspeed?

Docspeed combines OCR, layout analysis, extraction, grounding, async jobs, and document QA behind a hosted customer API. The API is designed for products that need reliable document automation and reviewer-visible source evidence.

Use it when raw OCR is not enough: AP invoice extraction, tax invoice processing, table review, schema-driven extraction, and multi-document question answering.

Documentation Index

For AI agents and internal tooling, a compact documentation index is available at /docs/llms.txt. The full text bundle is available at /docs/llms-full.txt.

Core Workflows

• Parse documents into OCR-backed markdown and page structure.
• Generate schemas from instructions, one document, or a document set.
• Extract fields and line items with evidence IDs.
• Review table cells with source-linked grounding.
• Queue longer workloads as async jobs.
• Ask grounded questions across one or more documents.

The Basic Workflow

1. Upload a document with POST /api/v1/upload.
2. Parse it with POST /api/v1/parse.
3. Generate or provide a schema.
4. Extract with POST /v1/extract.
5. Poll GET /v1/jobs/{job_id} and fetch GET /v1/jobs/{job_id}/result for async work.

Recommended Path

• Quickstart: make the first end-to-end call.
• Authentication & API Keys: send bearer auth correctly.
• Schema Builder: create schemas for extraction.
• Extract: use grounded extraction response shapes.
• Grounding: connect output values to document overlays.
• API Reference: inspect request and response contracts.

Product Language

Use the same canonical terms across integrations, docs, and outbound material:

• AP invoice extraction
• schema builder
• grounded extraction
• async jobs
• multi-document QA