Why is there a delay? (and what's normal)

A small delay between your speaker talking and the translation reaching listeners is completely normal. It’s a sign Glossa is doing its job carefully rather than rushing.

What’s normal

Captions appear a few seconds after the words are spoken, and the spoken translation lands roughly 7 to 10 seconds behind the speaker. Here’s why: the technology itself adds only 1 to 3 seconds, but Glossa waits for a sentence to finish so it has the full meaning, translates it, and then plays it back. Translating mid-sentence would be faster but far less accurate, so Glossa chooses accuracy.

This is by design

Because of that, there’s intentionally no “skip to live” button. Catching up by skipping would mean cutting off translated speech, and we’d rather your listeners get the complete, accurate message a few seconds later than a broken one instantly. When the audio does fall behind, Glossa speeds playback up for a while to catch up on its own.

We made a short video explaining the delay, which is handy to share with anyone who asks: https://www.youtube.com/watch?v=TE0vTHyNAH0

If your delay is much longer than that, or it keeps growing through the service, see Fixing a long or growing delay.