We count:

• The number of words in your documents.
• Which our ML models process for each insight in your chosen project type.

Each insight has an ML model that searches the text in your document, e.g. there is an ML model to look for Start Dates, another to search for Indemnities. There is also a small ML model that classifies documents when they are first uploaded, and this uses a flat rate of 250 ‘words’ regardless of how many words your document has.

Example: A user creates a project that has 10 insights, then uploads 20 documents with a total of 100,000 words.

First, 5,000 ‘words’ are consumed for uploading and classifying the documents (250 ‘words’ x 20 documents).

The user sends all 20 documents for review. This requires 1,000,000 credits (10 insight models x 100,000 words).

The total is therefore 1,005,000 ‘words’ to review these 20 documents.