ElevenLabs explains SSML tags for pauses, pronunciation, and pacing in text to speech

ElevenLabs says SSML is a W3C-rooted markup language that controls pauses, emphasis, pronunciation, and pacing in synthesized speech.

Its older models, including Multilingual v2, support the break tag for pauses up to three seconds, while Flash v2 and Turbo v2 also support the phoneme tag for English pronunciation.

Eleven v3 and Eleven v4 do not support SSML and instead use bracketed Audio Tags such as [whispering] and [sighs].