Start free. Scale as you grow. No hidden fees.
Perfect for testing and small projects
Documents go through Mistral OCR first ($0.001/page) for text extraction, then Claude for intelligent field mapping. This is 10x cheaper than sending raw images to vision models.
Free plans stop at the limit. Paid plans continue with per-extraction overage charges ($0.03 for Starter, $0.02 for Growth).
Yes! Upload 1-3 sample documents and our AI will automatically detect the fields to extract. Or define your schema manually with our builder.
Documents are processed in real-time and stored in encrypted Supabase storage. OCR results are cached by file hash for performance. You can delete your data at any time.