Rossum’s data extraction billing can be complex, but understanding which documents or pages are included in the billing is essential to managing your costs. Major aspect of our subscription is the volume allowance for document processing.

The fundamental principle of Rossum’s data extraction billing is that we charge for pages processed by our platform. This means that we count each page that you import into Rossum. Additionally, the billing mode is determined by the data capture configuration.

In this article, we’ll cover what you need to know about Rossum’s data extraction billing, including:

• Which documents or pages are included in the billing
• Which documents or pages are not included in the billing
• How billing is applied to split documents
• How to limit data extraction to the first N pages

Which documents or pages are included in the billing?

A document or page is billed when it is processed by Rossum’s AI Engine, which occurs when the document goes from the importing state to the review, confirmed, or exported state for the first time, depending on the configured automation level. 

For per-page billing, we charge for all document pages, regardless of which pages contain captured data. However, this only applies to the configured maximum number of pages per document processed by the AI Engine (first 32 pages by default).

Documents and pages included in the billing:

• Newly uploaded documents (up to the limit of pages processed by the engine): if the queue has a limit of 32 pages for data extraction and an uploaded document has 50 pages, only pages 1-32 will be included in the billing.
  
• Duplicates: currently, duplicate documents are processed by the AI Engine, and you will be charged for such documents.
  
• Documents that are automatically rejected based on a set of rules that can be set in Rossum.
  
• Documents moved to a queue with a higher limit of pages for extraction: if a document has 50 pages and is in a queue that has a 32-page limit for data extraction, initially only 32 pages will be included in the billing. Once the document is moved to another queue with a higher limit additional pages will be added to the billing.

Which documents or pages are not included in the billing?

While Rossum charges for most documents and pages processed by its AI Engine, there are exceptions to keep in mind to help manage your costs effectively. These include:

• Documents in the failed_import state.
  
• Deleted documents picked up by the “Delete recommendations” feature: the document will only be included in the billing if you decide not to delete it.
  
• Documents moved to another queue with the same pages for extraction limit: if you upload document to a queue that has a limit of 10 pages for data extraction and then you move it to another queue that has the same limit, you will be charged only for 10 pages in total.
  
• Documents moved to a queue with a lower limit of pages for extraction. 
  
• Documents moved between tabs (statuses) within the same queue: for example, changing from “Postponed” to “To Review”.
  
• Pages above the limit set for the queue: if the queue has a limit of 32 pages for data extraction and the uploaded document has 50 pages, only pages 1-32 will be included in the billing.

Document splitting

If you upload a document with many pages and then split it into several records, you won’t be charged twice. Here are some examples to help explain the billing rules:

Original document length	Split into	Pages deleted in the process	Amount of pages included in the billing
42 pages	2 documents (20 pages and 22 pages)	0	42 pages
42 pages	2 documents (20 pages and 20 pages)	2	40 pages

Limiting data extraction to first N pages

It is possible to configure the amount of pages included in the initial data extraction. You can specify the limit using the rir_params attribute of the queue object (using our API). For example, if you want to limit extraction to only the first page, you can set the effective_page_count parameter to 1.

It’s important to note that the default limit for a new queue is set to 32 pages, so if you don’t change it, Rossum will extract data from up to 32 pages and charge you accordingly. The extraction limit can be set to a number that is lower or higher than 32.

By limiting data extraction to the first N pages, you can save on processing costs and improve efficiency, especially for large documents where you may only need data from the first few pages.

If you need help with setting up the limit, please contact us – support@rossum.ai.