Web scraping & data APIs · Netherlands
Extract structured JSON from invoices, emails and PDFs with one API call
Ahrefs Domain Rating set in type 25 Oct 2026, reset 25 Oct 2026.
| Metric | Value |
|---|---|
| Domain Rating | 3 +1 (up) since last check |
| Trust Flow | 1 no change since last check |
| Citation Flow | 18 no change since last check |
| Referring domains | 8 no change since last check |
| External backlinks | 41 no change since last check |
Ahrefs Domain Rating set in type 25 Oct 2026, reset 25 Oct 2026. Lead topic: Recreation (1). Majestic set 22 Oct 2026, reset 22 Oct 2026. Link data typeset by Majestic. Trust Flow, Citation Flow and Topical Trust Flow are trademarks of Majestic-12 Ltd.
| Domain age | 6 months |
|---|---|
| Registered | |
| Expires | |
| Registrar | Cloudflare, Inc |
| DNSSEC | Not signed |
A document extraction API that converts unstructured text like invoices, emails, and PDFs into validated JSON matching your schema, with confidence scores for
Refractr is a structured data extraction API that transforms any document or payload into fields matching your own JSON schema. It processes invoices, emails, PDFs, and scanned documents using a custom-trained extraction model (not an LLM wrapper), returning typed fields with confidence scores in approximately 500 milliseconds per document. The API handles multiple document formats including PDF, DOCX, JPG, PNG, TIFF, WEBP, BMP, and TXT without requiring schema compilation or deployment. Every account includes 100 free extractions daily; additional volume uses prepaid credits at a flat rate of one credit per extraction, with scanned documents or photos requiring two credits due to OCR processing.
Pricing model: Freemium
Free tier: 100 free extractions per day. Pay as you go: €1 = 200 credits, one credit per extraction, top up from €5. Scans and photos with OCR: 2 credits.
No. Refractr runs a custom-trained, optimized extraction model, not a passthrough to OpenAI, Anthropic, or any general-purpose LLM. It is purpose-built for filling JSON schemas from raw text, which makes it faster than prompt-based approaches.
Fields declared with typed placeholders (like "date") are guaranteed to return that type or null, never an empty string or mistyped value. The guarantee covers the response shape; validate values against your own rules and enforce required fields in your schema validator.
Not unless you choose to. By default, documents and results are kept for 30 days for debugging, then deleted. You can opt in to help train the model, which increases your free allowance from 20 to 100 extractions per day. Your data is never sold or shared.
One credit per successful extraction, flat regardless of document or schema size. Failed extractions are free. Every account gets 100 free extractions per day, refreshed daily. For higher volume, top up any EUR amount via Stripe.
No. Pass the schema on every call. There is no compile step, no warm-up, and no per-schema deployment. You can change fields between requests with no penalty.
No review of Refractr set in type yet. Set the first.
Reviews are free to set. None are paid for. Read the proofs
Readers' notes · Reprints · Share the page · Your company in print? Claim it · Report a misprint
Refractr has a free listing. The owner can unlock the full profile with a paid launch for $5:
Browser-based tool to view public TikTok profiles, posts, engagement metrics and download videos anonymously without login or app installation.
ShoppingScraperReal-time e-commerce price monitoring API across 40+ countries
Reliable web data for AI
Can we set one analytics cookie to count which pages readers flip to? Nothing is set unless you say yes. Privacy