ABBYY
ABBYY ships as two separate packages, because they are two different products with different deployment models. Pick based on where the recognition runs.
| Package | Runs | Use it when |
|---|---|---|
| ABBYY Cloud | ABBYY's hosted service | You want OCR without installing anything. Submit an image, poll for the task, read the result. |
| ABBYY FineReader SDK | On the robot's machine | Documents cannot leave your network, or you need on-device camera OCR and classification. |
Both cover the same broad ground — extracting text from images and PDFs, reading barcodes, and processing structured documents such as invoices, receipts, business cards and MRZ travel documents.
Which to choose
ABBYY Cloud is the simpler of the two. It is task-based: you submit an image, the
service processes it asynchronously, and you poll until it is done. That asynchrony is
visible in the node list — Submit Image, Get Task Status, Wait Task, Delete Task —
so expect a flow with a wait in it rather than a single blocking call.
FineReader SDK processes locally and synchronously, which makes it the answer when documents must not leave the network, or when latency matters. It needs the SDK installed and licensed on the machine the robot runs on.
Related
- Document Processor — general document handling
- Tesseract — open-source OCR, no licence required
- Google Document AI and Google Vision — cloud alternatives