Azure AI Speech brings the Custom Avatar and Custom Video authoring experiences into the Microsoft Foundry portal in general availability. Approved customers can upload an actor's video recording, fine-tune a custom video avatar, and generate synthetic talking-head videos through the same Foundry project that runs their text-to-speech and Voice Live workloads. The portal also supports Photo Avatar, where a single reference image plus the Microsoft VASA model produces a head-only avatar. Customers can pair an avatar with a custom neural voice, a personal voice, or the new voice-sync-for-avatar option, then call the result through batch or real-time synthesis APIs. Custom Avatar remains a Limited Access feature gated by responsible AI review and an intake form.