Azure Document Intelligence - Can't get High Resolution to Run (SDK Example doesn't work).

Question

Azure Document Intelligence - Can't get High Resolution to Run (SDK Example doesn't work).

Timothy E Hu 20

.NET(v4.0)/Quickstarts/Samples/Sample_AddOnCapabilities_ExtractHighResolution.cs

// Specify DocumentAnalysisFeature.OcrHighResolution as the analysis features

        var features = new List<DocumentAnalysisFeature> { DocumentAnalysisFeature.OcrHighResolution };

        PromptRequestingDocumentIntelligenceService();

        var operation = await client.AnalyzeDocumentAsync(WaitUntil.Completed, "prebuilt-read", content, features: features);

        AnalyzeResult result = operation.Value;

        PromptGettingResponseFromDocumentIntelligenceService();
```I tried streaming content of URL but can't get it to work.  

I continually get a 404 error using the CODE from your SDK examples. 

           DocumentIntelligenceClient client = new DocumentIntelligenceClient(new Uri("https://psridocumentintelligence.cognitiveservices.azure.com/"), credential);

            //var client = new DocumentIntelligenceClient(new Uri(this.docIntelligenceEndPoint), new AzureKeyCredential(this.docIntelligenceApiKey));

            var content = new AnalyzeDocumentContent();

           

            if (!string.IsNullOrWhiteSpace(path))

            {

                #region Option 1: Analyze a sample document file from local file system.

                var docFileBytes = File.ReadAllBytes(path);

                content.Base64Source = BinaryData.FromBytes(docFileBytes);

                #endregion

            }

            else if (!string.IsNullOrWhiteSpace(url))

            {

                #region Option 2: Analyze a sample document file from url.

                var uriSource = new Uri(url);

                content.UrlSource = uriSource;

                #endregion

            }

            //  Specify DocumentAnalysisFeature.OcrHighResolution as the analysis features

            var features = new List<DocumentAnalysisFeature> { DocumentAnalysisFeature.OcrHighResolution };

            PromptRequestingDocumentIntelligenceService();

            var operation = await client.AnalyzeDocumentAsync(WaitUntil.Completed, "prebuilt-read", content, features: features);

            AnalyzeResult result = operation.Value;

![User's image](/api/attachments/891633e5-4a57-4c30-986b-8741a3569e3e?platform=QnA)

Ravada Shivaprasad 1,115 Reputation points Microsoft External Staff Moderator

2025-08-19T21:47:57.33+00:00

Hi Timothy E Hu

Microsoft's Document Intelligence service provides robust document analysis capabilities through its pre-built models, with the "prebuilt-read" model specifically designed for extracting text from various document types including PDFs, images, and office documents. The service runs at a higher resolution than Azure AI Vision Read and includes support for extracting both print and handwritten text, detecting paragraphs, text lines, words, locations, and languages.

The implementation process begins with proper initialization of the DocumentIntelligenceClient using Azure credentials and endpoint configuration. Once initialized, developers can utilize the AnalyzeDocumentAsync method to process documents from URLs, specifying the "prebuilt-read" model and enabling high-resolution OCR through the DocumentAnalysisFeature parameter. The service processes the document asynchronously, allowing for efficient handling of large documents while maintaining optimal system resources.

The Document Intelligence service supports a comprehensive range of document formats, including PDF, JPEG, PNG, BMP, TIFF, HEIF, DOCX, XLSX, PPTX, and HTML files. For optimal results, ensure that the document URL is publicly accessible and that the file format is among the supported types. The high-resolution OCR feature is particularly valuable for documents containing small text or complex layouts, where standard OCR might struggle to maintain accuracy.

When implementing URL-based document analysis, it's essential to consider several key factors. First, verify that your Document Intelligence instance is properly configured in the Azure portal. Ensure your API credentials are valid and that the document URL is accessible from the service. The service will automatically detect text orientation and layout, making it suitable for processing documents regardless of their physical characteristics.

For production implementations, consider implementing proper error handling and logging mechanisms. The service provides detailed status information throughout the analysis process, allowing for robust monitoring and debugging capabilities. By leveraging these features along with the high-resolution OCR option, developers can build reliable document processing solutions that maintain high accuracy across various document types and qualities.

Reference links : Document Intelligence read model , Document Intelligence add-on capabilities

Hope it Helps!

Thank you
Ravada Shivaprasad 1,115 Reputation points Microsoft External Staff Moderator

2025-08-21T00:59:08.7533333+00:00

Hi Timothy E Hu

Did you get any chance to check the response.

Thank you!
Timothy E Hu 20 Reputation points

2025-08-21T02:24:20.6466667+00:00

Yes the response provided information I already had. I can't get the example to work at all. Using valid end point and key - problem with call giving 404 error. The following is the call - I have tried local content / changing the content as URL / changing content to Binary but nothing works. Just need to get PDF run in High Resolution.Operation<AnalyzeResult> operation = await client.AnalyzeDocumentAsync(WaitUntil.Completed, "prebuilt-read", content);
Ravada Shivaprasad 1,115 Reputation points Microsoft External Staff Moderator

2025-08-28T17:30:39.73+00:00
Hi Timothy E Hu

Sorry for Delayed response

You're encountering a 404 Not Found error when using the AnalyzeDocumentAsync method with the "prebuilt-read" model in Azure Form Recognizer, and this issue can stem from several common misconfigurations. First, ensure that your endpoint is correctly formatted—typically it should look like https://<your-resource-name>.cognitiveservices.azure.com/. Any deviation from this format, such as missing slashes or incorrect subdomains, can cause the API to fail.

Next, verify that you're using the correct model ID. "prebuilt-read" is valid, but it must be used with the newer DocumentAnalysisClient from the Azure.AI.DocumentIntelligence namespace, not the older Form Recognizer SDK. Also, confirm that your SDK version supports this method; older versions may not be compatible.

When passing content, make sure you're using the correct format. For binary content, use BinaryData.FromStream(pdfStream), and for URLs, use AnalyzeDocumentContent with a UrlSource. Additionally, check that your Azure resource is properly deployed in a supported region and that the Document Intelligence service is active. Some users have reported that their resource didn’t appear in the Document Intelligence section, which led to 404 errors.

Finally, ensure your client is initialized correctly with both the endpoint and the API key

var client = new DocumentAnalysisClient( new Uri(endpoint), new AzureKeyCredential(apiKey) );

Hope it helps!

Thank you

Your answer

Ravada Shivaprasad 1,115 Reputation points Microsoft External Staff Moderator

2025-08-21T00:59:08.7533333+00:00

Hi Timothy E Hu

Did you get any chance to check the response.

Thank you!
Timothy E Hu 20 Reputation points

2025-08-21T02:24:20.6466667+00:00

Yes the response provided information I already had. I can't get the example to work at all. Using valid end point and key - problem with call giving 404 error. The following is the call - I have tried local content / changing the content as URL / changing content to Binary but nothing works. Just need to get PDF run in High Resolution.Operation<AnalyzeResult> operation = await client.AnalyzeDocumentAsync(WaitUntil.Completed, "prebuilt-read", content);
Ravada Shivaprasad 1,115 Reputation points Microsoft External Staff Moderator

2025-08-28T17:30:39.73+00:00

Hi Timothy E Hu

Sorry for Delayed response

You're encountering a 404 Not Found error when using the AnalyzeDocumentAsync method with the "prebuilt-read" model in Azure Form Recognizer, and this issue can stem from several common misconfigurations. First, ensure that your endpoint is correctly formatted—typically it should look like https://<your-resource-name>.cognitiveservices.azure.com/. Any deviation from this format, such as missing slashes or incorrect subdomains, can cause the API to fail.

Next, verify that you're using the correct model ID. "prebuilt-read" is valid, but it must be used with the newer DocumentAnalysisClient from the Azure.AI.DocumentIntelligence namespace, not the older Form Recognizer SDK. Also, confirm that your SDK version supports this method; older versions may not be compatible.

When passing content, make sure you're using the correct format. For binary content, use BinaryData.FromStream(pdfStream), and for URLs, use AnalyzeDocumentContent with a UrlSource. Additionally, check that your Azure resource is properly deployed in a supported region and that the Document Intelligence service is active. Some users have reported that their resource didn’t appear in the Document Intelligence section, which led to 404 errors.

Finally, ensure your client is initialized correctly with both the endpoint and the API key

var client = new DocumentAnalysisClient( new Uri(endpoint), new AzureKeyCredential(apiKey) );

Hope it helps!

Thank you

Share via

Azure Document Intelligence - Can't get High Resolution to Run (SDK Example doesn't work).

Your answer