Certificate OCR API
Extract names, titles, issuers and dates from diplomas, training certificates and awards for verification.
結果
変換されたMarkdownやJSONがここに表示されます。
What makes certificates hard to read
- Script and decorative fonts
- Seals, signatures and borders over text
- No two issuers use the same layout
What you can pull out
certificate fields
{
"recipient": "Ahmed Rahman",
"title": "Bachelor of Engineering (Civil)",
"issuer": "University of Leeds",
"issue_date": "2024-07-15",
"certificate_id": "UL-2024-118734",
"grade": "First Class Honours"
}Best models for certificates
1Gemini FlashGeneral vision model with strong handwriting and layout understanding at a low price.
google/gemini-flash3 credits / page2GPTGeneral vision model. Good at messy scans and following extraction instructions.openai/gpt11 credits / page3ClaudeGeneral vision model with careful reading of long, dense documents.anthropic/claude14 credits / pageCall it from your code
curl https://api.ocrrouter.ai/v1/ocr \
-H "Authorization: Bearer $OCRROUTER_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "google/gemini-flash",
"file_url": "https://example.com/invoice.pdf",
"format": "json"
}'