Terrapin DocExtract Beta

Automatically extract structured information from financial documents.
Simply upload a pdf, list the fields you want to extract, and our tool will use AI to find them in the document.
You can then download the results as a CSV file.


What kind of documents is this tool made for?

This tool is document-agnostic, though it was built primarily to extract structured data from security prospectuses.

What's the maximum document size you support?

You may submit up to 5 files at once, with at most 300kb worth of text each. This corresponds roughly to 40-page documents. If you want to be able to submit more files at once or submit larger files, get in touch.

How accurate is this tool?

Although we try our best to make the model output consistent results, you may need to tweak the field names in order to get the results you are looking for. We also note that the model is not deterministic, so different runs may result in different outputs.

Does this tool use GPT or a similar model?


Should I submit private or sensitive data to this tool?

No. We may keep some of the data submitted by users, so we recommend you only submit documents that are publicly available and not proprietary.