
ElevenLabs v3 Text-to-Speech API by ELEVENLABS
ElevenLabs v3 Text-to-Speech model. High-quality speech generation from text prompts.
ElevenLabs V3 Text-to-Speech
ElevenLabs V3 Text-to-Speech is ElevenLabs' latest flagship speech synthesis model, delivering highly expressive, natural-sounding audio from text. With improved emotional range, multilingual fluency, and a diverse library of 21 built-in voices, V3 sets a new standard for AI voice generation.
Why Choose This?
-
Expressive voice quality V3 produces natural, human-like speech with nuanced emotion and intonation.
-
Large voice library Choose from 21 built-in voices covering a range of genders, ages, and speaking styles.
-
Flexible text normalization Control how numbers, abbreviations, and symbols are spoken with auto, on, or off modes.
-
Adjustable stability Fine-tune voice consistency versus expressiveness to match your content.
-
Long-form support Input up to 5,000 characters per request for articles, narrations, and scripts.
Parameters
| Parameter | Required | Description |
|---|---|---|
| text | Yes | The text to convert to speech. Maximum 5,000 characters |
| voice | No | Voice to use (default: Bella). See voice list below |
| stability | No | Voice stability from 0 (expressive) to 1 (consistent), default: 0.5 |
| apply_text_normalization | No | Text normalization mode: auto (default), on, or off |
Available Voices
| Voice | Voice | Voice | Voice | Voice |
|---|---|---|---|---|
| Bella | Roger | Sarah | Laura | Charlie |
| George | Callum | River | Harry | Liam |
| Alice | Matilda | Will | Jessica | Eric |
| Chris | Brian | Daniel | Lily | Adam |
| Bill |
How to Use
- Write your text — provide the content to be spoken, up to 5,000 characters.
- Select a voice — choose from the 21 built-in voices based on gender, tone, and style.
- Adjust stability (optional) — lower values give more expressive delivery; higher values give consistent, neutral tone.
- Configure text normalization (optional) — use
autofor most cases,onto always expand numbers/abbreviations,offto skip normalization. - Run — submit the request and retrieve the generated audio URL.
Best Use Cases
- Content Narration — Convert articles, blogs, and documents into natural audio.
- Audiobook Production — Generate long-form narration with consistent voice quality.
- Voiceover & Media — Create professional voiceovers for videos, ads, and presentations.
- Multilingual Applications — Deliver localized speech in 30+ languages.
- Conversational AI — Power chatbot and virtual assistant voice responses.
- Accessibility Tools — Build screen readers and assistive listening applications.
Pro Tips
- Use
BellaorSarahfor warm, professional narration; useRogerorGeorgefor authoritative or conversational tones. - Set
stabilityto0.3–0.4for storytelling and emotional content; use0.7–0.9for news reading or technical narration. - Set
apply_text_normalizationtoonwhen your text contains numbers, currencies, or abbreviations that must be spoken out in full. - Break very long content into chunks under 5,000 characters and stitch the audio outputs for seamless long-form production.
Pricing
| Billing Standard | Price |
|---|---|
| Per 1,000 characters | $0.003 |
Notes
modelandtextare required fields.- Maximum input length is 5,000 characters per request.
- Task status values:
created,processing,completed,timeout,failed. - Audio output URLs are returned in
data.outputsonce status iscompleted.
Related Models
- ElevenLabs V2 Text-to-Speech — Previous generation multilingual speech synthesis model.
- ElevenLabs Voice Cloning — Clone a custom voice from a reference audio sample.



