Voice & music
Eleven v4 - ElevenLabs v4
ElevenLabs’ most expressive voice model, plus a Turbo version for live agents.
Voice & music
ElevenLabs’ most expressive voice model, plus a Turbo version for live agents.
Product insights
Natural, acted-sounding speech from one model, with a real-time variant for conversations.
Most TTS voices sound flat on long scripts or are too slow for live conversations.
Creators, audiobook and ad producers, and developers building voice agents.
Free plan with 10,000 credits a month; paid plans start at $6/month. v4 and v4 Turbo share the same credit pricing.
Eleven v4 is the fourth generation of ElevenLabs’ speech model, the one people mean by “ElevenLabs v4”. It comes in two versions. Eleven v4 is tuned for produced audio such as voiceovers, audiobooks, and ads, with context stitching that keeps pacing and tone steady across long scripts of up to 10,000 characters. Eleven v4 Turbo is tuned for live calls and AI agents, streaming speech in about 150 ms. Both work with all 17,500+ library voices and, unlike v3, with Professional Voice Clones.
Choose from 17,500+ library voices, design a new one, or clone your own.
Add audio tags like [laughs], [whispers], or [door slams] where you want them.
Render audio in the web app, or call eleven_v4 / eleven_v4_turbo through the API.
It is Eleven v4, ElevenLabs’ text-to-speech model released on September 28, 2026, together with Eleven v4 Turbo for real-time use.
Eleven v4 aims for the best quality in produced audio. v4 Turbo keeps the same expressive range but streams in about 150 ms, so it suits live calls and AI agents.
v4 is more consistent on long scripts, sequences audio tags better, adds a real-time Turbo version, and brings back Professional Voice Clones, which v3 did not support.
You can try it on the free plan with 10,000 credits a month. Paid plans start at $6/month.
Official website snapshot
elevenlabs.io
Eleven v4 (searched as “ElevenLabs v4”) is the text-to-speech model ElevenLabs released on September 28, 2026. It reads a script the way a voice actor would, in 90+ languages, with audio tags like [laughs] and [whispers]. Eleven v4 Turbo brings the same voices to live AI agents at about 150 ms to first speech.