Refractr
Fastest structured JSON extraction from documents
Details
- Target Audience
- DevelopersData EngineersBusiness Analysts
- Pricing
- Freemium
- Platforms
- API
Discovery signals
How AI and people discover Refractr on PeerPush
About Refractr
Refractr is a document extraction API that converts unstructured data into validated JSON. It allows you to transform invoices, emails, and PDFs and more into structured data using your own custom schemas in less than 500 milliseconds per document. The service streamlines the process of data capture(OCR) and extraction with a single API call, ensuring that extracted information adheres to specific validation requirements for reliable downstream processing, while giving confidence scores unlike standard LLM's. This reduces the time and money spent on manual extraction review.
Product Video
Watch a video demo of Refractr.
Screenshots
Reviews (4)
Average 5.0 out of 5
Based on 4 reviews
Better option than textract or docAI and surely LLM's
Still in alpha release but already very fast and accurate! Will give it a proper try when you officially launch
Comments (2)
How are you able to reach an extraction time of a few hundred milliseconds? Is this a model like Jev?
@Semtech Due to having a tiny model made for one specific job, we can stay accurate. We profile where time gets wasted during inference and fix it. The inference time itself is only about 250 ms, the rest is API and networking overhead.
@sepp We can expect another 2x speedup soon due to better hardware
Structured JSON straight from invoices is a big time saver. Curious how it handles multi-page scanned PDFs with tables.
@nckbrz GIve it a try! You have 100 free extractions a day.