Share via

No MRZ extracted

Claudio Pizzillo 20 Reputation points
2026-03-19T08:18:38+00:00

Hi, using APi 2024-11-30 (GA) with prebuilt-idDocument, in the new european identity cards, the MRZ code is not extracted. Tested also in Document Intelligence Studio with various cards from many countries. ..

Azure Document Intelligence in Foundry Tools
0 comments
Answer accepted by question author
  1. SAI JAGADEESH KUDIPUDI 2,210 Reputation points Microsoft External Staff Moderator
    2026-04-14T17:33:19.1066667+00:00

    Hi Claudio Pizzillo,
    thank you for bringing this to our attention and for validating the behavior using both the API version 2024‑11‑30 (GA) and Document Intelligence Studio across multiple European ID cards.

    Based on the current documented capabilities of Azure AI Document Intelligence (prebuilt‑idDocument model), this behavior is expected.

    At present, while European national ID cards are supported as document types, the MRZ (Machine Readable Zone) present on newer EU identity card formats is not exposed as a dedicated structured field in the model output when using the prebuilt-idDocument model in v4.0 (2024‑11‑30 GA).

    This explains why MRZ data is not returned even though it may be visually present on the document and processed successfully in Document Intelligence Studio.

    As an immediate and Microsoft‑recommended approach, if the MRZ is visible on the ID card, you may:

    • Process the same document using a general text extraction model such as Read or Layout
    • Extract the MRZ content as raw OCR text
    • Apply custom parsing logic within your application to interpret MRZ fields

    This method allows you to retrieve MRZ information reliably for downstream processing, even though native structured MRZ output is currently not available for EU ID cards in the prebuilt-idDocument model.

    Additionally, you may consider training a custom extraction model tailored to your specific EU ID formats if MRZ extraction is a critical business requirement, as Azure Document Intelligence supports custom model training for documents with varying templates and layouts.
    Conclusion

    To summarize, the MRZ not being extracted from newer European national ID cards when using the prebuilt‑idDocument model is aligned with the current model capabilities rather than a configuration issue.

    While native MRZ extraction for these ID formats is not yet supported as structured output, using the Read/Layout OCR model combined with custom MRZ parsing logic can serve as an effective workaround to meet business requirements until future model enhancements expand native support.
    Microsoft Reference Links

    Hope this helps. If you have any follow-up questions, please let me know. I would be happy to help.

    Please do not forget to "Accept Answer" and "up-vote" wherever the information provided helps you, as this can be beneficial to other community members.

    0 comments

1 additional answer

Sort by: Most helpful
Most helpful Newest Oldest
  1. Q&A Assist
    2026-03-19T08:39:44.1833333+00:00
    AI generated content. This question contains an answer created with AI from Q&A Assist. Learn more

    With API version 2024-11-30 (GA) and the prebuilt-idDocument model, European ID cards and driver licenses are supported as document types, but the documentation does not list MRZ (machine-readable zone) lines on EU ID cards as dedicated, structured fields.

    The behavior observed—no MRZ extracted for new European ID cards, even when tested directly in Document Intelligence Studio across multiple countries—is consistent with the currently documented capabilities. The ID model has expanded support for EU IDs (as noted in the February 2024 updates), but this expansion is described in terms of document and field coverage in general, not MRZ-specific extraction.

    Given the available information, MRZ on European ID cards is not exposed as a specific field in the prebuilt-idDocument model output in v4.0 (2024-11-30 GA).

    To work with MRZ data in this scenario:

    1. Confirm current behavior
      • Use Document Intelligence Studio with Model ID: prebuilt-idDocument and inspect the full JSON result to verify that MRZ is not present as a named field. This matches current documentation.
    2. Use generic text extraction as a workaround
      • If the MRZ is visually present on the card, process the same image with a general text/structure model (for example, a layout/read-style model from Document Intelligence) and then parse the MRZ in custom code.
    3. Track feature evolution and submit feedback
      • The ID document model is actively evolving (for example, “expanded field support for European Union identification cards and driver licenses” in the February 2024 release notes). Monitor the ID document model and What’s new pages for any future announcement that adds explicit MRZ extraction for EU IDs.
      • If MRZ extraction for European ID cards is a key requirement, submit it as a feature request through Azure support/feedback channels so it can be considered for future updates.

    References:

Your answer

Answers can be marked as 'Accepted' by the question author and 'Recommended' by moderators, which helps users know the answer solved the author's problem.