Inquiry on Migration to Image Analysis v3.2 or v4.0

Question

Hello Azure Team,

We are currently using Computer Vision Image Analysis v2.0 and are planning to migrate to either v3.2 or v4.0 (linhttps://learn.microsoft.com/en-us/azure/ai-services/computer-vision/overview-image-analysis?tabs=4-0). Upon comparing Image Analysis v3.2 and v4.0, we noticed that the following features are missing from v4.0:

• Brands
• Faces
• Image type
• Color scheme
• Categories (Landmarks, Celebrities)
• Adult content

For our use cases, we require all visual features, including those currently missing from v4.0. We would appreciate your advice and feedback on the following questions:

1. Which of these missing features are scheduled to be added in v4.0?
2. What is the timeline for adding these features to v4.0 (in GA release)?
3. For Faces and Adult/Safe content features, is there a plan to add them as part of Image Analysis v4.0, or do we need to use their dedicated services (i.e., Azure AI Face service/Azure AI Content Safety)?
4. Since Image Analysis v3.0 and v3.1 are retiring on September 13, 2026, is there a plan/timeline to retire v3.2? Will there be 3.2+ versions (3.3, 3.4), or should everyone eventually migrate to v4.0?
5. For our use case, what do you recommend? Should we migrate to v3.2 and stay in v3.2+, or eventually migrate to v4.0?
6. In Image Analysis v2.0 and v3.2, we have Image Description, and in v4.0, we have image captions. As noted, the version 4.0 image captioning model is more advanced but is only available in certain regions (link). Is there a plan/timeline to have this feature in all regions?
7. Is there a plan to have Image Description in v4.0, or will we only have image captions/dense captions?

Many thanks in advance!

Answer 1

Hello @Ashkan Izadpanah , Thanks for using Microsoft Q&A Platform.

We really appreciate your interest in learning more about the availability and timelines of the features. Most of your questions seem to be about our roadmap or timelines. Unfortunately, at this point, we don't have any specific ETA, or roadmap details we can share.

Please visit what's news page for the latest information and updates: https://learn.microsoft.com/en-us/azure/ai-services/computer-vision/whats-new

1. Which of these missing features are scheduled to be added in v4.0? Unfortunately, we don't have any information to share with you at this time.
2. What is the timeline for adding these features to v4.0 (in GA release)? We don't have an ETA.
3. For Faces and Adult/Safe content features, is there a plan to add them as part of Image Analysis v4.0, or do we need to use their dedicated services (i.e., Azure AI Face service/Azure AI Content Safety)? I would suggest you use the Azure AI Face service for face detection and analysis. For adult and safe content features, the Azure AI Content Safety service is recommended
4. Since Image Analysis v3.0 and v3.1 are retiring on September 13, 2026, is there a plan/timeline to retire v3.2? Will there be 3.2+ versions (3.3, 3.4), or should everyone eventually migrate to v4.0? Since the documentation does not mention any retirement announcement now for V3.2, it is advisable to use this. However, the latest version is recommended.
5. For our use case, what do you recommend? Should we migrate to v3.2 and stay in v3.2+, or eventually migrate to v4.0? Consider your requirements and use-case while choosing the version. However, general recommendation is to go for the latest versions.
6. In Image Analysis v2.0 and v3.2, we have Image Description, and in v4.0, we have image captions. As noted, the version 4.0 image captioning model is more advanced but is only available in certain regions (link). Is there a plan/timeline to have this feature in all regions?
7. Is there a plan to have Image Description in v4.0, or will we only have image captions/dense captions? Unfortunately, no ETA regarding these.

I hope this helps. Thank you again for your time and patience throughout this issue.

Regards,

Vasavi

Please remember to "Accept Answer" if any answer/reply helped, so that others in the community facing similar issues can easily find the solution.