Known Issue - “Latest” Voices Failing to Generate Audio

Date: July 13, 2026

Description: Some text-to-speech voices labeled “Latest” may fail to generate audio. After processing briefly, users may see an error asking them to refresh the page and try again.

In some cases, replacing an existing voiceover with an affected voice may remove the original audio and leave the audio track blank.

Workaround: Split longer scripts into two or more shorter sections and generate each section separately. Users can also select a comparable voice that is not labeled “Latest.”

Scope: Azure text-to-speech voices labeled “Latest” across multiple languages. The issue may affect the standalone Text to Speech page, voiceovers created within videos, and AI Avatar generation.

STATUS UPDATE

[July 31, 2026, 12:00 AM PT]: Our Engineering team is working on a backend update to address this issue. The update is currently scheduled for release on August 19, 2026.

[July 15, 2026, 12:00 AM PT]: Engineering determined that the affected voices may time out while processing. The issue has been reported to the voice provider for further investigation.

[July 13, 2026, 5:30 AM PT]: Engineering is investigating reports that some voices labeled “Latest” fail to generate text-to-speech audio.

If you are still having issues or have questions, please contact support@vyond.com.