Extrait Kbis OCR API
Automate the extraction of critical business information from French Extrait Kbis documents. Get structured data like SIREN, SIRET, company name, legal form, and executives.
Extract SIREN and SIRET reliably
Automatically find and validate the core identification numbers of French businesses.
Identify legal representatives
Extract the names, roles, and details of presidents, directors, and other key executives.
Capture company status
Get the exact legal form (SAS, SARL, etc.), share capital, and registration dates.
Handle scanned copies
Process low-quality scans or photos of physical Kbis documents with high accuracy.
How it works
Get from a raw document to structured JSON in three simple steps.
Choose or Upload
Choose OR Upload your own document to generate an schema from it
Customize the Schema
curl -X POST "https://tiny-idp-api-338302005544.europe-southwest1.run.app/api/extractors/run/YOUR_EXTRACTOR_ID" \
-H "x-api-key: YOUR_API_KEY" \
-F "files=@/path/to/your/document.jpg"Start Extracting
Start by uploading one of your Kbis documents
Our IA will extract the fields from your Kbis document. Then, you can customize to your needs.
Playground: Create from a document
Upload any document and our AI will design a custom extractor tailored to it.
Input Document
Upload your own
PDF, PNG, JPG
Output & Integration
Ready for production?
Get an API key to start extracting data from your documents in production.
Start from scratch
Know exactly what you need? Define your own JSON schema manually.
Why automate French Extrait Kbis documents?
The Extrait Kbis is the standard proof of a French company’s legal existence and key registry facts. In onboarding, KYC, vendor due diligence, and credit workflows, teams repeatedly open PDFs or scans to copy SIREN, legal form, and signatories—steps that delay decisions and invite typos.
Automating Kbis extraction gives you consistent, auditable company data in seconds: faster partner verification, fewer manual checks, and easier integration with CRM, risk, and compliance tools. It is especially useful for banks, insurers, marketplaces, and B2B platforms that verify many French entities.
OCR tuned for registry layouts also handles stamped extracts and imperfect scans, so you can process real documents rather than only perfect digital copies.
90% Faster
Reduce processing time from minutes to seconds.
Higher Accuracy
Eliminate manual entry errors and typos.
What fields can be extracted from a French Extrait Kbis?
You can extract core registry identifiers—SIREN and SIRET—the company’s legal name (dénomination), legal form (e.g. SAS, SARL), registered office address, share capital, activity description or NAF/APE code when present, dates of incorporation or extract, and company status. You can also capture governance data: names and roles of legal representatives (président, gérant, etc.) and other entries as shown on the document. With a custom schema you can normalize how names, addresses, and identifiers are split for your KYC or master-data rules.
Commonly extracted fields include:
Simple, Transparent Pricing
No hidden fees. No monthly minimums. Pay only for what you extract.
Usage-Based
Simple pay-as-you-go pricing. No monthly commitment.
Enterprise
Tailored pricing for high-volume scenarios. Get SLA guarantees, on-premise deployment, and dedicated support — reach out and we'll put together a plan that fits your scale.
All prices exclude VAT. Volume discounts apply automatically.
Enterprise-grade Compliance & Security
We take data privacy seriously. Tiny IDP is built from the ground up to meet the strictest European data protection standards.
Zero Data Retention
We don't store your documents, images, or predictions. Data is processed in-memory and immediately discarded.
GDPR Compliant
Full compliance with European data protection regulations (GDPR) for your peace of mind.
EU-Based Infrastructure
All data is processed and hosted exclusively in secure European data centers.
Do you need a custom OCR?
We support custom extractors! Define your own fields, rules, and logic to extract data from any type of document.