ElevenLabs ships Eleven v4: audio tags direct emotion and voice clones are indistinguishable
thealexbanks · x · 2026-09-29
ElevenLabs has released Eleven v4, its most expressive voice model yet — the author says he can no longer tell his cloned voice from the real one.
Key features:
- Audio tags: insert tags to direct emotion, pacing and reactions; v4 follows stacked tags more reliably than v3
- Voice clones: hold their identity across regenerations
- Target duration: fit audio to a set length
- Available in 90+ languages
Real-world use cases include audio newsletter overviews in your own cloned voice, video ad hooks from proven email lines, one voice across all dubbing languages, and fixing a single audiobook sentence without re-recording the chapter. Workflow: pick Eleven v4 in Text to Speech, choose a voice, and direct delivery via tags.
Related event: ElevenLabs Launches Eleven v4, Tops TTS Leaderboards at a Premium Price(22 posts)→
More from Multimodal
- User Shares Coral Reef Creation Labeled 'Qwen 4 Max (xhigh)', an Unannounced Model Name — intellectronica · 2026-11-16
- Reddit megathread crowdsources the best open-source audio models: ASR, TTS, and text-to-music — rm-rf-rm · 2026-11-16
- Animating AI-generated 2D game art with FLUX.2 klein: masked redraws work but remain a hack — emilrueh · 2026-11-16
- Hand-editing 0.2% of weights plus 100 CPU-minutes turns an English TTS model into an 18MB Sanskrit one — kastnerkyle · 2026-11-16
- Zyphra open-sources textnorm tool, lifting TTS structured-data accuracy by up to 22 points — kastnerkyle · 2026-11-16
- AI music video tells a racing driver's comeback story — VictoryVisualsStudio · 2026-11-16