Azure OpenAI is prematurely cutting off support for Sora 2 too early

Azure OpenAI Foundry currently plans to prematurely end support for Sora 2 on June 6, 2026, even though OpenAI’s timeline indicates support should continue until September 24, 2026. It feels way too early and inconsistent with the broader timeline.

Azure also contradicts themselves previously announced on X that they would extend support. I created a Microsoft Q&A post asking them to honor that extension through September 24 (unfortunately I can’t share the link here, but it’s pretty easy to find on Google azure sora 2 june 6). I would really appreciate it if others could reply on the thread and ask Microsoft to align Sora 2 Foundry’s retirement date with OpenAI’s timeline.

If anyone has ideas for other effective ways to raise visibility with the right Microsoft or Azure OpenAI teams, please share them. This seems like the kind of issue that needs more attention before the earlier cutoff becomes final.

This is especially important for teams like ours. As part of the Microsoft for Startups program, we receive credits that can only be used on Microsoft’s platform. Moving directly to OpenAI endpoints would create a cost burden we currently can’t absorb.

Sora 2 is burning compute - which OAI cannot afford. This is well known. Looks like Microsoft feels the same way.

The real problem is not whether Sora 2 is expensive to run, but that Azure appears to be walking back its own commitment. They said on X that Foundry would support it for “6 months,” yet the announced June 6 retirement date does not line up with that at all.

Keep in mind that this is an OpenAI developer forum - It has nothing to do with Microsoft or control over what Microsoft does.

That’s fair, and I understand. I only raised it here because Azure OpenAI Foundry is a major way many developers access OpenAI models.

Yes it is! Thank you for sharing it here!

I updated your original post.

Free service provided by the moderators.