Document Reader SDK 9.8 — September 7, 2026
IMPORTANT changes
- Added new neural networks for more accurate detection and recognition of Arabic text for Saudi Arabia.
- Added new neural networks for more accurate detection and recognition of embossed text.
- Fixed an issue with Complete Server-Side Verification where RFID chip data received from Mobile SDK 9.4 and earlier led to an incorrect CA result. We strongly recommend updating Web API to this version to ensure correct reprocessing results. If the Web API and Mobile SDK versions match, reprocessing works correctly.
Core SDK
Document Detection and Crop
- Updated the server-side neural network for the document detection and crop with greater accuracy.
- The
doublePageSpreadproperty continues to work in single-frame processing mode, contrary to what was stated in version 7.1, to support cases where not all pages are required. It is enabled by default in this mode. If disabled, only one page is extracted.
OCR
- Added new neural networks for more accurate detection and recognition of Arabic text for Saudi Arabia.
- Added new neural networks for more accurate detection and recognition of embossed text.
MRZ
- Updated parsers:
- Aruba ID Card
- Cyprus ePassport
- Cyprus Residence Permit
- Fixed an issue with CAN detection for Estonian ID cards in scenarios without document type identification.
Barcode
- Added new parsers:
- Nepal DL
- Syrian Arab Republic Temporary Registration Certificate
- Viet Nam DL
- Updated parsers:
- Colombia Residence Permit
- Jordan Digital ID Card
- India Digital Certificate of Registration
- Updated the parser for QR codes generated by a Spanish government application—MiDNI.
Text data parsing and validation
- Fixed an issue with parsing and transliteration of Argentine surnames and given names.
- Fixed an issue with parsing and transliteration of French surname and given names.
Authenticity
- Updated the neural network for the electronic device replay check with greater accuracy.
- Updated the neural network for the document occlusion check with greater accuracy.
- Updated the neural network for the hologram detection check with greater accuracy for Moldova ID cards.
- Fixed an issue where the rear page had an incorrect page index, which led to the incorrect electronic device result “Invalid input data”.
Other
- Fixed an issue with Complete Server-Side Verification where RFID chip data received from Mobile SDK 9.4 and earlier led to an incorrect CA result. We strongly recommend updating Web API to this version to ensure correct reprocessing results. If the Web API and Mobile SDK versions match, reprocessing works correctly.
- Optimized memory usage.
- Improved code quality with minor issues fixed.
- Updated dependencies to the latest versions to address security vulnerabilities.
Core RFID SDK
- Optimized memory usage.
Enums
- Added new
eRFID_ManualChipDetectionModeenum member:ddmRestartChip
- Added new
eRFID_Commandsenum members:RFID_Command_PICC_OpenRFID_Command_PICC_CloseRFID_Command_PICC_APDURFID_Command_PICC_Config
- Added new
eRFID_Application_Typeenum members:at_AppletRootat_UserDefined
- Added new
eRFID_Password_Typeenum member:pptPIN_Local
Hardware Support
- Fixed an issue with redundant repeated capture under UV light on Thales readers.
Mobile API
Android
- Improved document capture in the FullAuth scenario on Samsung S26 when working with Mobile Document Readers.
- Fixed an issue with video recording in the FullAuth scenario when 1920x1440 resolution was used.
iOS
- Fixed an issue where Bluetooth permission was required when it was not necessary.
Web Components
- Added hidden text labels for screen readers and other assistive technologies to conform to WCAG recommendations (
aria-labelattributes). - Fixed an issue with incorrect default camera selection on Huawei devices.
- Improved frame capture performance.
- Fixed an issue with preview size on iOS 27 in Safari.
- Updated dependencies to the latest versions to address security vulnerabilities.
UI Components
- Added support for the following languages: Arabic, Armenian, Bangla, Croatian, Czech, Danish, Dutch, Finnish, French, Greek, Hebrew, Hindi, Hungarian, Indonesian, Italian, Japanese, Korean, Kyrgyz, Malay, Norwegian, Polish, Portuguese, Romanian, Slovak, Spanish, Swedish, Thai, Turkish, Ukrainian, Uzbek, and Vietnamese.
- Added the ability to configure whether to show the download button for images.
- Updated dependencies to the latest versions to address security vulnerabilities.
- Other minor improvements.
Web Service API
- Updated dependencies to the latest versions to address security vulnerabilities.
Clients
- Fixed an issue with setting the
tenantandenvproperties in the Python client. - Updated dependencies to the latest versions to address security vulnerabilities.
Demo Website
- Updated the Document Reader Web Component presentation to full-screen mode on mobile devices.
Database
- Updated multiple document templates to include OVI, MLI, Contact chip, Holograms, Extended MRZ, Extended OCR, Photo embedding type, Patterns checks where appropriate.
- Includes 254 countries and territories, 16881 documents.
- Added 74 new documents.
For details, see the Supported documents list.