OpenVidu 2.25.0: AWS and Vosk support for Speech to Text

OpenVidu 2.25.0 is out, and as co-founders of the project we're glad to share what's new. Full details are in the release notes; the highlights:

• Speech to Text engine support expanded: AWS Transcribe joins the existing Azure option, and Vosk is now available too — an open-source engine that needs no cloud-provider key and adds no extra cloud cost, a real alternative for anyone who'd rather not depend on Azure or AWS for transcription.
• A new OPENVIDU_PRO_DOCKER_REGISTRIES configuration property, letting you point Media Nodes at private Docker registries for the kurento-media-server and speech-to-text-service images instead of only public ones.
• A fix for a recordingActivity directive typo in OpenVidu Components.

The Speech to Text change is the one worth a closer look if you're already using (or considering) live transcription: Vosk in particular removes the cloud-provider dependency entirely, which matters for anyone who wants transcription without adding another external service to their bill or their compliance scope. The full changelog is in the GitHub release.