| Version | Change log |
| JAWS 2025.2412.50 Dec 18, 2024 |
Enhancements in JAWS 2025.2412.50 (December 2024): Convenient OCR Updates: JAWS 2025 offers improvements to the Convenient OCR (Optical Character Recognition) feature. These include a new method for recognizing the text of images based on the current context as well as the ability to perform OCR continuously in the background. All-in-one Keystroke: The new keystroke reduces the effort of using Convenient OCR by using only, INSERT+SPACEBAR, O, ENTER, to recognize the text from an image based on your current application and context. This eliminates the need to remember multiple keystrokes constantly. For instance: If you are in File Explorer, it recognizes the currently selected image file. If you are in Adobe Reader, it recognizes the currently open PDF document. If you are playing a YouTube video, it activates Background OCR and places the Invisible cursor in the video area so you can access any displayed text as it is recognized. If you are in a Teams or Zoom meeting where content is being shared, such as PowerPoint slides, it activates Background OCR and places the Invisible cursor in the shared content area so you can view the recognized text as it updates. In any other applications, it recognizes the current window that has focus. Background OCR: To help identify and read text from applications that are otherwise inaccessible, JAWS 2025 can now continuously perform OCR in the background for the active window. This ensures users can seamlessly access and interact with text content on their screens, improving overall accessibility and user experience. To start Background OCR, use the layered command, INSERT+SPACEBAR, O, B. To review the OCR results, turn on the JAWS or Invisible cursor. Continue navigating with the JAWS or Invisible cursor to read the recognized text as it updates. Braille display users can also position themselves on a line and watch it update in real-time as new text is recognized. Tip: Pressing INSERT+SPACEBAR, O, CTRL+B also starts background |
| JAWS 2025.2410.10 Nov 1, 2024 |
Automatic Language Switching: The default for JAWS automatic language detection has been changed back to detect languages Based on Markup of Documents and Web Pages, since the standard for specifying languages in documents and web pages is to use the appropriate language tags. This will remain the default for the initial 2025 release. If you encounter a situation where JAWS does not switch the voice to a specific language even if an appropriate voice is available, open Settings Center, expand the Text Processing group, expand the Language processing subgroup, and change the Detect Language setting to use the new option, Based on AI predictions. Thank you to everyone who provided feedback on this new feature. Aria and Web: JAWS now says “on” or “off” when interacting with switch elements on web pages. When encountering interactive elements on a web page that use the Separator ARIA role and JAWS Forms Mode is set to Auto or Semi-Auto, Forms Mode activates when focus moves to these elements. When using Auto Forms Mode and you navigate in and out of an edit field using UP or DOWN ARROW, addressed an issue where the prompt was announced when exiting the field instead of the next item gaining focus. Using table navigation commands (ALT+CTRL+ARROWS) in some web-based applications no longer causes the browser’s menu bar to activate. In Forms Mode, reduced speech verbosity when navigating ARIA menus by eliminating additional “menu” announcements for subsequent menu items. If a table on a web page supports changing the sort order of a column, JAWS now properly recognizes the new sort order and speaks and displays it in Braille. Labels for edit fields on web pages are now displayed correctly in Braille. In certain instances, radio buttons no longer appear as having multiple labels. If Mouse Echo is enabled, JAWS reads open browser tabs when hovering over them with a mouse. Improved JAWS interaction with selecting items in Angular tree grids. When reading a PDF fi |
| JAWS 2024.2409.2 Sep 11, 2024 |
Picture Smart AI Updates: This update offers even more improvements to Picture Smart AI thanks to the great feedback we continue to receive from users. These include a new method for describing an image based on the current context as well as the ability to begin asking for specific details the first time an image is submitted. All-in-one Keystroke: The new keystroke reduces the effort of using Picture Smart AI by using only, INSERT+SPACEBAR, P, ENTER, to get a description based on your current application and context. This eliminates the need to remember multiple keystrokes constantly. For instance: If someone is presenting in Teams, it describes the current slide. If you are on a web page, it describes the current image or control. If you are in File Explorer, it describes the currently selected file. If you are in PowerPoint, it describes the entire screen. Ask a Question Before Submitting a Picture: In situations where you already have some basic knowledge about an image, you can now include ALT with any Picture Smart command to immediately ask a question about specific details. For example, suppose you were shopping on Amazon for a product like an Multicooker or Air Fryer. You already have a basic idea of what the product is, but you may want to know upfront if the product contains physical buttons or a touch activated display. You could press INSERT+SPACEBAR, P, ALT+C, type your question followed by ENTER, and Picture Smart AI attempts to answer the question about the selected image instead of just providing a general description. Once the results containing the response are displayed, you can still use the Ask questions about this image link to ask follow-up questions if needed. Feedback for Cropped Images: After running Picture Smart AI, you will now be alerted to any potential problems with an image that may result in poor descriptions. For example, if your browser window is not maximized, images may be clipped, leading to inaccurate descriptions. In this c |
| JAWS 14.0.5005 Aug 13, 2013 |
