ElevenLabs Eleven v4
ElevenLabs' most expressive model: emotion and delivery via inline audio tags, 90+ languages
What it's the best tool for
- Inline audio tags control emotion: [laughs], [whispers], [sighs], [excited]
- More than 90 languages, picked from a list or detected automatically
- 21 ready voices: George, Sarah, Adam, Alice, Daniel, Lily and more
- Fine settings: stability, similarity, style exaggeration, speaker boost, text normalization
- Up to 10,000 characters per request
- Export to MP3 and WAV
When to reach for something else
- No speech speed control
- No voice cloning on NetRoom, only the ready voices are available
- Up to 10,000 characters per request, longer scripts have to be split
- Billing follows the number of characters, not audio duration
How ElevenLabs Eleven v4 responds
ElevenLabs Eleven v4
Four scenarios where it pays for itself
More about ElevenLabs Eleven v4
ElevenLabs Eleven v4 Online — Expressive AI Voice
ElevenLabs Eleven v4 is the most expressive text-to-speech model from ElevenLabs. It does not just read a script, it performs it: inline audio tags control the delivery, and the voice-over can be set to any of 90+ languages. It runs in the browser on NetRoom, with nothing to install.
Audio tags: direction inside the script
Put a cue in square brackets before a line or in the middle of it — [laughs], [whispers], [sighs], [excited] — and the voice changes manner: it laughs, drops to a whisper, sighs, or speaks with energy. The tags themselves are never spoken, they only shape the delivery. Flat reading turns into a scene with a mood.
90+ languages and 21 ready voices
The model speaks more than 90 languages. In the editor you can pick a language from a list (Russian, English, Ukrainian, German, French, Spanish, Japanese, Korean, Chinese and more, 18 options in total) or leave it on auto-detect. The library has 21 ready voices: George, Sarah, Adam, Alice, Daniel, Lily and others.
Fine control over delivery
Every generation offers voice stability, similarity to the voice, style exaggeration, speaker boost, and a text normalization mode of auto, on, or off. Normalization decides how numbers and abbreviations are read. This model has no speed control: rhythm comes from punctuation, pauses, and tags.
Who it is for
Audiobook and podcast authors get lifelike narration with emotion. Game developers and animators get character lines with the right mood. Marketers get ads and promo voice-overs. Localization teams voice one script in dozens of languages.
How to voice a script
1. Sign up on NetRoom and top up your balance. 2. Open the Sound section and pick ElevenLabs Eleven v4. 3. Paste up to 10,000 characters and place audio tags. 4. Choose a voice and language, and tune the delivery if you like. 5. Download the result as MP3 or WAV.
Pricing
No subscription: billing follows the number of characters you voice. The current rate is shown on this page. Voice cloning is not available on NetRoom, the ready voices are used.
More ElevenLabs models
For faster generation with the same tags, see Eleven v4 Turbo. For a speed control and steady reading of long scripts, see Multilingual v2. The previous generation with audio tags is Eleven v3.
Voice your first script — start on NetRoom.
What changed ElevenLabs Eleven v4
- + Added voice model ElevenLabs Eleven v4 (ElevenLabs).
Use ElevenLabs Eleven v4 via the API
The same engine, straight from your code: one key and one balance for text, images, video and sound. Pay only for the requests you make.
curl https://netroom.ai/api/v1/generations \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model": "elevenlabs/eleven-v4", "input": {"prompt": "A cinematic mountain sunrise"}}'The model id is already in the example. The full parameter reference and prices live in GET /api/v1/models and in the docs.
Try ElevenLabs Eleven v4
right now
Free access to basic models. No card, no obligations.