![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(262.40000000000003, 25%, 35%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3EA%3C/text%3E%3C/svg%3E)
Hi Microsoft team, is it still needed to use an external OCR tool like Poppler? I am also trying to create an agent able to run OCR automatically on some scanned files but it completely fails.
thanks
Enabling OCR for JPEG Scan Documents in Copilot Studio Agent
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(137.6, 65%, 23%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3EWC%3C/text%3E%3C/svg%3E)
W C H Bagya Perera
5
Reputation points
Dear Team,
Currently using Copilot Studio with the GPT-4 model to build an agent that connects to a SharePoint knowledge base. The knowledge base contains a mix of JPEG scanned documents uploaded by users. However, the agent is currently unable to read or respond to queries based on the content in these JPEG files, as it appears that OCR (Optical Character Recognition) is not enabled or supported by default.
I would like to know:
- Is it possible to enable OCR for the Copilot Studio agent (GPT-4) so it can extract and understand text from JPEG scan documents stored in SharePoint?
- If yes, what are the recommended steps or configurations to enable OCR support within the agent or its connected SharePoint data source?
Your guidance on how to proceed would be greatly appreciated.
Microsoft Copilot | Other
-
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(160, 48%, 25%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3ESM%3C/text%3E%3C/svg%3E)
Sayali-MSFT 5,771 Reputation points Microsoft External Staff Moderator2025-08-26T09:19:37.92+00:00 Hello @W C H Bagya Perera, thank you for bringing this issue to our attention. We will look into it and get back to you shortly.
Sign in to comment
2 answers
Sort by: Most helpful
-
Sayali-MSFT 5,771 Reputation points Microsoft External Staff Moderator
2025-08-27T09:15:17.34+00:00 Hello @W C H Bagya Perera,
Currently, Copilot Studio (GPT-4) does not natively support OCR (Optical Character Recognition) for extracting text from JPEG or other image files in SharePoint knowledge bases. The agent can only process text-based content by default.
You must extract text from images using an external OCR process and make that text available to the agent, either as metadata or via a custom extension.Reference Document-https://learn.microsoft.com/en-us/ai-builder/prebuilt-text-recognition
Thanks,
Sayali
*************************************************************************
If the response is helpful, please click on "upvote" button. You can share your feedback via Microsoft Copilot Developer Feedback link. Click here to escalate.
-
Sayali-MSFT 5,771 Reputation points Microsoft External Staff Moderator
2025-09-01T10:05:28.2733333+00:00 Hello @W C H Bagya Perera, Could you please confirm if your issue has resolved with provided suggestions or still looking for any help?
Sign in to comment -