ElevenLabs launches Eleven v4 and v4 Turbo with 150 ms time-to-first-speech

ElevenLabs released Eleven v4 (quality-first) and v4 Turbo (speed-first) on September 28, 2026. Turbo hits ~150 ms median time-to-first-speech in streaming mode, supports 90+ languages, and clones a voice from a 10-second sample.

• Eleven v4 Turbo: ~150 ms time-to-first-speech, ~100 ms inference
• 90+ languages supported, up from 70 in v3
• Voice cloning from a 10-second audio sample
• Follows inline tags like [softly] and [laughter] in text