Drop in a PDF, image, or DOCX — a single file or a batch. Each upload can be up to ≤10MB and ≤15 pages.
UPLOAD · EXTRACT · EXPORT
Extract structured data from any document.
Upload PDFs, images, or DOCX and get clean, typed JSON — powered by OCR and leading LLMs, through a dashboard or a REST API.
WHY DOCR
Built on leading models, designed for production.
Authenticate with a scoped key, post a document, and get structured JSON back. Authorization: Bearer docr_sk_… → typed Invoice JSON, every time.
Two models, one API
Claude Opus 4.8 for the highest quality, Gemini 3 Flash for the fastest turnaround — pick per request.
12 document types built in
Invoices, tax forms, IDs, statements and more work out of the box — and you can define custom types of your own.
Secure by default
HMAC-signed webhooks, IP whitelisting, and scoped API keys so your document pipeline stays locked to your infrastructure.
HOW IT WORKS
Upload. Extract. Export.
Three steps from a raw file to clean, typed data — in the dashboard or over the API.
Step 1
Upload
Step 2
Extract
OCR reads the page and an LLM maps it to structured fields. Choose Claude Opus 4.8 for quality or Gemini 3 Flash for speed.
Step 3
Export
Get clean, typed JSON in the dashboard or via the API — and a signed webhook the moment the job finishes.
DOCUMENT TYPES
12 built-in types — plus your own.
Common business and tax documents are ready to extract on day one. Need something else? Define a custom type with your own fields.
INVOICE
Invoice
Line items, totals, tax, and vendor details as typed fields.
W-2
W-2
Wages, withholdings, and employer information.
DRIVER'S LICENSE
Driver's License
Name, address, license number, and expiry.
PASSPORT
Passport
MRZ fields, holder details, and document number.
BANK STATEMENT
Bank Statement
Balances, transactions, and account metadata.
UTILITY BILL
Utility Bill
Service address, billing period, and amount due.
PAY STUB
Pay Stub
Gross, net, deductions, and pay period.
PURCHASE RECEIPT
Purchase Receipt
Merchant, line items, tax, and total.
1040
1040
U.S. individual income tax return fields.
1099-*
1099-*
The 1099 family of information returns.
TAX FORMS
More tax forms
Additional built-in tax document types, ready to extract.
CUSTOM
Need something else?
Define custom types with your own typed fields →
CORE FEATURES
Everything you need to turn documents into data.
Batch extraction
Process one file or a whole batch in a single run.
Choose your model
Claude Opus 4.8 for the highest quality, Gemini 3 Flash for the fastest response.
Extraction history
Browse past extractions, view the JSON, and download results.
Custom document types
Define your own document types with typed fields.
PDF · JPG · JPEG · PNG · BMP · WEBP · DOCX
Any common format
Files up to ≤10MB and ≤15 pages per upload.
Dashboard or API
Work in the dashboard or wire extraction into your stack via REST.
BUILT FOR DEVELOPERS
A REST API built for document pipelines.
Authenticate with `Authorization: Bearer docr_sk_…`, post a document, and get structured JSON back. Subscribe to `extraction.completed` and `extraction.failed` webhooks to drive the rest of your pipeline.
curl https://api.docr.dev/v1/extractions \ -H "Authorization: Bearer docr_sk_live_…" \ -F "file=@invoice.pdf" \ -F "type=invoice" \ -F "model=claude-opus-4-8"
SECURITY & CONTROL
Signed webhooks. IP allowlists. Your data, controlled.
Every webhook carries X-docr-Signature: t=…,v1=… (HMAC-SHA256) and retries on failure, so you can verify authenticity before you trust a payload.
X-DOCR-SIGNATURE
Signed webhooks
HMAC-SHA256 signatures on every event, with automatic retries until your endpoint acknowledges.
IPV4 · CIDR
IP whitelist
Restrict API access to known IPv4 ranges or CIDR blocks — everything else gets a 403.
CLERK ORGANIZATIONS
Team management
Roles and shared access through Clerk Organizations.
Integrated with the tools you already run.
Turn your documents into data today.
Start on the free plan — 50 pages/mo, no credit card required.