Document Extraction API: Invoices, DDT and Orders to Structured JSON
Add AI document data extraction to your own product or workflow. Data Alchemy exposes its Intelligent Document Processing engine through a REST API and webhooks: send invoices, delivery notes (DDT), orders or contracts and get back structured, validated JSON — 99.8% accuracy with Claude, around 12 seconds for a one-page document (average time with Claude Sonnet), ready to post into your ERP.
The REST API and webhooks are included in every licence: Basic, Business and Enterprise.
An IDP API that turns documents into structured data
The document extraction API is the programmatic side of Data Alchemy's Intelligent Document Processing platform. Instead of clicking through a web app, your systems submit documents and receive validated, structured data in return — so you can embed invoice, DDT and order extraction directly into your own software, automate your accounts-payable pipeline, or feed any ERP or CRM without manual re-keying. For REST endpoints, webhook payloads and the output data schema, see the developer API documentation. The same contract is published as an OpenAPI 3.1 specification — import it into Postman, a client generator or an AI coding assistant and scaffold the integration without transcribing anything.
Still choosing? Our comparison of the best document processing APIs for invoice and receipt extraction puts this API next to LLM-native and cloud alternatives, with public pricing and sources.
One request, structured data back
Authenticate with your API key, POST the document as multipart/form-data with its document_type, and read the validated fields from a GET on the document id — or let the HMAC-signed webhook push them to you when processing completes.
curl -X POST https://api.data-alchemy.ai/v1/documents \
-H "Authorization: Bearer $DATA_ALCHEMY_API_KEY" \
-H "Content-Type: multipart/form-data" \
-F "file=@invoice.pdf" \
-F "document_type=invoice"{
"id": "doc_8a7f2c91",
"status": "processing",
"document_type": "invoice",
"created_at": "2026-06-04T09:12:33Z",
"webhook_url": "https://yourapp.example.com/hooks/data-alchemy"
}curl https://api.data-alchemy.ai/v1/documents/doc_8a7f2c91 \
-H "Authorization: Bearer $DATA_ALCHEMY_API_KEY"
{
"document_type": "invoice",
"header": {
"supplier": {
"name": "Rossi Forniture S.r.l.",
"vat_number": "IT01234567890"
},
"invoice_number": "2026/00417",
"issue_date": "2026-05-28",
"currency": "EUR"
},
"line_items": [
{
"sku": "ART-0042",
"description": "Cartone 30x20x15",
"quantity": 12,
"unit_price": 8.50,
"total": 102.00,
"vat_rate": 22
}
],
"totals": { "net": 102.00, "vat": 22.44, "gross": 124.44 },
"validation": { "status": "validated", "erp_match": true }
}The same payload reaches your endpoint on the document.processed event when you prefer an asynchronous pipeline, signed with HMAC SHA-256 in the X-Data-Alchemy-Signature header. For the full endpoint reference, every accepted document_type, the complete output schema and error handling, see the developer API documentation.
Everything you need to integrate invoice, DDT and order extraction
REST API for document data extraction
Submit a document and retrieve its extracted, validated data as structured JSON — header fields, line items, totals, taxes and references — ready to map onto your ERP.
HMAC-signed webhooks for asynchronous pipelines
Build asynchronous pipelines: Data Alchemy notifies your endpoint when each document has been processed, so you never poll for results.
AI agent: extracted data posted into your ERP or CRM
Prefer not to build the integration yourself? The AI agent, installed at your premises, connects to Data Alchemy over websocket and APIs and posts the data into your ERP or CRM on a database, CSV, JSON or TXT files, or APIs, with your ERP credentials kept in your perimeter. It is a separately quoted add-on on the Business licence and included in Enterprise.
Email acquisition with the on-premise Mail service
The Mail service, installed at your premises, manages multiple business mailboxes from a graphical interface and sends Data Alchemy the emails matching a model — zero-code integration, with emails and credentials kept in your perimeter.
Structured JSON validated against your ERP master data
Every document returns clean, typed JSON validated against your master data, so downstream systems consume reliable data instead of raw text.
A dedicated LLM per document model
Each document type is processed by the single engine chosen for its model (Claude by Anthropic, GPT by OpenAI, Data Alchemy AI, Data Alchemy's proprietary model, or a compatible model through the European providers Scaleway and Regolo.ai); with Claude it reaches 99.8% accuracy (no accuracy figure is published for the other engines), with no templates; reading takes around 12 seconds for a one-page document (average time with Claude Sonnet or higher).
From document to structured data in four steps
Submit
Send a document to the REST API or, from the Business licence upwards, let the on-premise Mail service send it from your business mailboxes.
Extract & classify
A dedicated LLM classifies the document and extracts every field — no layout training, no templates.
Validate
The extracted data is checked against your ERP master data in real time, flagging anomalies before they propagate.
Deliver
Receive structured JSON via API response or webhook, then post it into SAP, Zucchetti, TeamSystem or any other system.
Explore the platform behind the API
SAP, Zucchetti, TeamSystem
How extracted data is written into your management system through the AI agent or the REST API and webhooks.
Learn more →ReconciliationLLM invoice reconciliation
Automated 3-way matching between invoices, orders and delivery notes.
Learn more →IDP platformDocument management software
The AI layer that reads, understands and extracts what a DMS only stores.
Learn more →Frequently asked questions about the document extraction API
What is the Data Alchemy document extraction API?
It is the programmatic interface to Data Alchemy's Intelligent Document Processing (IDP) engine. Instead of using the web app, your systems submit documents — invoices, delivery notes (DDT), purchase orders, contracts, price lists — and receive back structured, validated data ready to write into your ERP or CRM.
Which integration methods are available?
There are four modes. The REST API with signed webhooks, for developers. The REST API is included in every licence (Basic, Business, Enterprise). The AI agent, installed at your premises, which connects to Data Alchemy over websocket and APIs and posts the data into your ERP on the destination configured from a graphical interface: a database, CSV, JSON or TXT files, or APIs (a separately quoted add-on on the Business licence, included in Enterprise). The native Salesforce connector with Order Dispatcher, which creates draft orders in the CRM (Enterprise licence). The Mail service, installed at your premises, which captures documents from business mailboxes (from the Business licence).
What format does the extracted data come back in?
Extracted fields are returned as structured JSON — header data, line items, totals, taxes and document references — already validated against your ERP master data, so it can be mapped directly onto your system of record.
How accurate and fast is the extraction?
Data Alchemy pairs each document model with a single AI engine: Claude (Anthropic), the engine we recommend by default; GPT (OpenAI); Data Alchemy AI, Data Alchemy's proprietary model, running on its own servers and GPUs in Italy; or a compatible LLM through the European inference providers Scaleway and Seeweb Regolo.ai, on the customer's own subscription. With Claude it reaches 99.8% accuracy (no accuracy figure is published for the other engines) with no templates and no per-layout training; with Claude Sonnet or higher, around 12 seconds for a one-page document and around 18 seconds for two pages (average times).
Which ERPs and systems can I write the data into?
The API is system-agnostic: with the REST API and webhooks you write the JSON into SAP, Zucchetti, TeamSystem or any other ERP, CRM or internal application yourself. If you'd rather not build the integration, the AI agent, installed at your premises, posts the data into your ERP via a database, CSV, JSON or TXT files, or APIs.
Build with the document extraction API
The REST API is included in every licence. Tell us about your use case: we'll give you your API key and walk you through integration on your real documents — no commitment.
Request your API key