PDF analysis, table extraction, document classification, and merchant statement parsing.
Try endpoints directly in your browser with Swagger UI.
/docsClean, readable endpoint documentation.
/redocAI-friendly markdown format. Paste into any LLM context.
/api-reference.mdMachine-readable schema for SDK generation and AI agents.
/openapi.json| Method | Endpoint | Description | Auth |
|---|---|---|---|
| GET | /health | Health check | None |
| POST | /analyze | Upload PDF, extract text/tables/KV, apply parsing profiles | Any key |
| POST | /extract-tables | Extract tables from PDF (JSON, CSV, or XLSX) | Any key |
| POST | /classify | Classify or label a document (human + auto) | Any key |
| GET | /doc-types | List known document types | Any key |
| POST | /doc-types | Register a new document type | Master |
| GET | /classify/stats | Classification training progress | Any key |
| POST | /keys | Create a new sub API key | Master |
| GET | /keys | List all API keys with usage | Master |
| GET | /keys/{id}/usage | Detailed usage for a key | Master |
| DELETE | /keys/{id} | Revoke an API key | Master |
| POST | /pipedrive/sync | Sync files from won Pipedrive deals | Master |
| GET | /pipedrive/sync/status | Last sync status and counts | Master |
| GET | /pipedrive/deals | List synced deals and files | Master |
All endpoints except /health require a Bearer token:
Authorization: Bearer your_api_key_here
Master key: Full access. Create sub keys, manage Pipedrive sync, register doc types.
Sub keys: Scoped to specific endpoints, rate-limited, with per-key usage tracking (pages processed, requests, cache hits).
To integrate this API with an AI agent or tool:
/openapi.json for the full machine-readable spec/api-reference.md for an LLM-friendly markdown summary/.well-known/ai-plugin.json