Eleven v3
eleven_v3
The most expressive model. Supports 70+ languages. Requires more prompt engineering than our previous models.
Examples
Parameters
| Name | Type | Default | Constraints | Description |
|---|---|---|---|---|
| voice_id *声音id | text | — | — | |
| language_codeISO 639-1语言code | text | — | — | |
| similarity_boostSimilarity Boost | slider | 0.75 | 0 ~ 1 · step 0.05 | |
| stabilityStability | slider | 0.5 | 0 ~ 1 · step 0.05 |
Output fields
| Field | Type | Description |
|---|---|---|
| audio_url | string |
API
Call this model through one unified REST API. Get a key on the API Keys page.
cURL
# 1) Submit — returns { "task_uuid": "..." }
curl -X POST "https://www.namifusion.com/api/v1/marketplace/run/eleven_v3" \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"input": {
"voice_id": "JBFqnCBsd6RMkjVDRZzb",
"language_code": "en",
"similarity_boost": 0.8,
"stability": 0.55
}
}'
# 2) Poll until status is "completed", then read the output URLs
curl "https://www.namifusion.com/api/v1/marketplace/run/tasks/TASK_UUID" \
-H "Authorization: Bearer YOUR_API_KEY"Python
import time, requests
API_KEY = "YOUR_API_KEY"
HEADERS = {"Authorization": f"Bearer {API_KEY}"}
# 1) Submit
resp = requests.post(
"https://www.namifusion.com/api/v1/marketplace/run/eleven_v3",
headers=HEADERS,
json={
"input": {
"voice_id": "JBFqnCBsd6RMkjVDRZzb",
"language_code": "en",
"similarity_boost": 0.8,
"stability": 0.55
}
},
)
resp.raise_for_status() # 401/402/429/5xx stop here instead of polling a bad task
task = resp.json()
# 2) Poll until a terminal state (completed / failed / cancelled).
# This model is allowed up to 300s server-side.
deadline = time.time() + 360
while task.get("status") not in ("completed", "failed", "cancelled"):
if time.time() > deadline:
raise TimeoutError(f"still {task.get('status')} — keep the task_uuid and poll later")
time.sleep(3)
poll = requests.get(f"https://www.namifusion.com/api/v1/marketplace/run/tasks/{task['task_uuid']}", headers=HEADERS)
poll.raise_for_status()
task = poll.json()
print(task["status"], task.get("output"))JavaScript
const API_KEY = "YOUR_API_KEY";
const HEADERS = { Authorization: `Bearer ${API_KEY}` };
// 1) Submit
const resp = await fetch("https://www.namifusion.com/api/v1/marketplace/run/eleven_v3", {
method: "POST",
headers: { ...HEADERS, "Content-Type": "application/json" },
body: JSON.stringify({
"input": {
"voice_id": "JBFqnCBsd6RMkjVDRZzb",
"language_code": "en",
"similarity_boost": 0.8,
"stability": 0.55
}
}),
});
if (!resp.ok) throw new Error(`submit failed: ${resp.status} ${await resp.text()}`);
let task = await resp.json();
// 2) Poll until a terminal state (completed / failed / cancelled).
// This model is allowed up to 300s server-side.
const deadline = Date.now() + 360 * 1000;
while (!["completed", "failed", "cancelled"].includes(task.status)) {
if (Date.now() > deadline) throw new Error(`still ${task.status} — keep the task_uuid and poll later`);
await new Promise((r) => setTimeout(r, 3000));
const poll = await fetch(`https://www.namifusion.com/api/v1/marketplace/run/tasks/${task.task_uuid}`, { headers: HEADERS });
if (!poll.ok) throw new Error(`poll failed: ${poll.status}`);
task = await poll.json();
}
console.log(task.status, task.output);Documentation
Eleven V3
A high-naturalness text-to-speech model for fast, expressive voice generation with no cold starts.
Eleven V3 is ElevenLabs’ hosted text-to-speech model designed to turn written text into clear, lifelike, and expressive speech. It combines strong voice quality, low operational friction, no cold starts, and affordable usage-based pricing, making it a practical choice for production voice workflows.
🚀 Key Features
- High naturalness: Produces human-like intonation, pacing, and articulation for polished audio output.
- Expressive speech synthesis: Delivers clear pronunciation with lifelike emotional tone, suitable for narration and branded content.
- Controllable voice behavior: Fine-tune output with similarity and stability to balance realism, consistency, and speaker likeness.
- Speaker Boost support: Improves English text normalization, especially for numbers, times, and measurements.
- Multi-language capability: Supports multiple languages with strong overall intelligibility, while performing best in English.
- No cold starts: Hosted inference is ready to use for responsive, production-friendly generation.
🛠️ Technical Specifications
| Item | Details |
|---|---|
| Model architecture | Deep-learning text-to-speech(TTS)model |
| Model name | Eleven V3 |
| Input | Text string, voice_id string |
| Output | Audio file URL |
| Output format | MP3 |
| Maximum input length | 10,000 characters |
| Billing unit | 1 character = 1 token; charged per character |
| Voice options | Built-in voice library including Alice, George, Lily, Daniel, and more |
| Adjustable controls | similarity(0.0–1.0), stability(0.0–1.0), use_speaker_boost(boolean) |
| Pause control | Use <#x#> between words to control pause length(0.01–99.99s) |
| Language support | Multi-language text-to-speech; best performance in English |
| Latency | No fixed latency published; hosted deployment is positioned for no-cold-start responsiveness |
| Best practice | Split long passages into smaller segments for more stable rhythm and delivery |
Sample prompts
Note: This is a text-to-speech model, so it does not use Prompt / Negative Prompt in the image-generation sense. The examples below are sample input texts.
Welcome to our advanced text-to-speech system! Experience high-quality voice synthesis with natural pronunciation and clear articulation.The meeting will begin at 9:30 AM <#0.5#> and will cover sales performance, regional forecasts, and next quarter planning.Welcome to our product overview. Today, we’ll introduce a faster and more natural voice experience for digital content.
💰 Pricing
| Item | Price |
|---|---|
| Base price | $0.1 / 1,000 characters |
💡 Best Use Cases
- Voiceovers for digital content: Ideal for explainer videos, product demos, training content, and branded media.
- Podcasts and spoken media: Useful for summaries, narration, educational audio, and serialized content.
- Business audio automation: Suitable for announcements, internal training, spoken instructions, and customer-facing audio assets.
- Multilingual content production: A strong fit for global marketing, localized product messaging, and international media workflows.
🔗 Related Models
- Elevenlabs Dubbing: Better suited for dubbing and cross-language voice replacement workflows.
- Elevenlabs Eleven V3 Timing: More appropriate for use cases that require tighter timing or alignment control.
Related models
Eleven Turbo v2.5
Our high quality, low latency model in 32 languages. Best for developer use cases where speed matters and you need non-English languages.
Eleven Multilingual v2
Our most life-like, emotionally rich mode in 29 languages. Best for voice overs, audiobooks, post-production, or any other content creation needs
Eleven Flash v2.5
Our ultra low latency model in 32 languages. Ideal for conversational use cases.