PDF से डेटा निकालें
Gharauni PDF Parser API.
Turn scanned property cards into structured JSON in under 2 seconds.
Every loan origination system (LOS) we've talked to has the same bottleneck: villagers send their Gharauni as a PDF or photo, and a human has to type out the 13-digit ID, plot number, area, and coordinates. Our parser does it in 1.4 seconds with 98% accuracy. ₹5 per parse at scale. Free trial: 50 parses.
The data entry tax on rural lendingमानव की गलती का बोझ
When a villager applies for a Gharauni-backed loan, the bank receives a PDF or phone photo of the card. Today, that gets fed into a manual review queue. A clerk types out the 13-digit ID, plot number, owner name, area, and tehsil into the LOS. This process takes 4-8 minutes per application, has a 6% manual-error rate, and bottlenecks the entire rural lending pipeline. At 100 applications/day, that's 8 person-hours and ~6 typo errors that propagate downstream.
Structured extraction via REST APIREST API से डेटा
POST a PDF or image to our /api/parse endpoint. Receive structured JSON in 1.4 seconds median: { gharauniId, plotNumber, ownerName, area, district, tehsil, village, coordinates, issueDate, confidence }. Our model is trained on 50,000+ real SVAMITVA cards across UP, MH, MP, HR, KA. Handles low-quality phone photos, scanned PDFs, and printed copies.
Result: Cuts loan-application processing time by 80%. Saves ₹4-6 per loan in clerk costs.
What the API returns
OCR + layout understandingचित्र समझ
Reads structured data from any quality scan or photo. Trained on real SVAMITVA card formats across 5 states with regional script variants.
JSON outputJSON बाहर
Clean, validated JSON ready for your LOS. Includes confidence scores per field so you know when to flag for human review.
Private + ephemeralसुरक्षित
PDFs deleted from our servers after 24 hours. DPDP-compliant. We never train on your customer data. SOC 2 Type I in progress.
How to integrate
Sign up, get API key
Free trial with 50 parses, no credit card. Instant API key. Self-serve dashboard for usage tracking.
POST your PDF
Single endpoint: POST /v1/parse with multipart/form-data. Or send a public URL. cURL or any HTTP client.
Receive JSON
Median 1.4 seconds. Per-field confidence scores. Webhook callback option for async use.
Frequently Asked Questions · अक्सर पूछे जाने वाले सवाल
What's the accuracy rate?सटीकता कितनी है?
What's the accuracy rate?सटीकता कितनी है?
On clean PDFs from the SVAMITVA portal: 99.2% field-level accuracy. On phone photos of physical cards: 96.4%. Each response includes per-field confidence so you can flag low-confidence extractions for human review.
What's the pricing?दाम क्या है?
What's the pricing?दाम क्या है?
Free: 50 parses/month. Starter: ₹5/parse, capped at ₹5,000/month. Growth: ₹3/parse at 5k+/month. Enterprise: custom pricing at 50k+/month. Volume pricing email enterprise@gharauni.com.
Do you support languages other than Hindi/English?अन्य भाषाएँ?
Do you support languages other than Hindi/English?अन्य भाषाएँ?
Yes — Marathi, Telugu, Kannada, Haryanvi, Gujarati cards all supported in v2 (Aug 2025+). Bilingual cards parsed in both languages with cross-validation.
What rate limits should I expect?दर सीमा?
What rate limits should I expect?दर सीमा?
100 requests/minute on free tier. 1000/min on Starter+. Enterprise plans have custom limits up to 10k req/min with reserved capacity.
Is there an SDK?SDK?
Is there an SDK?SDK?
Python and Node.js SDKs in /parser/docs. Open-sourced under MIT. Java/Go on the roadmap.
How does this compare to AWS Textract or Google Document AI?AWS Textract से क्या अंतर?
How does this compare to AWS Textract or Google Document AI?AWS Textract से क्या अंतर?
General-purpose document AIs achieve 70-80% accuracy on SVAMITVA cards because they don't know the layout. We're trained specifically on this single document class. Net: higher accuracy, lower latency, 1/5 the price.
Stop typing 13-digit IDs by hand.मैनुअल डेटा एंट्री बंद करें।
50 free parses to try it on your own loan application backlog. No credit card. Cancel anytime.
Start free trial →