ElevenLabs Turbo v2.5

Fast, high-quality text-to-speech from ElevenLabs with extensive voice options.

Property	Value
Model ID	`elevenlabs/turbo-v2.5`
Providers	ElevenLabs, Replicate
Type	Text-to-speech

Basic Usage

import { compose, generateAudio, audioModel } from "@synthome/sdk";

const execution = await compose(
  generateAudio({
    model: audioModel("elevenlabs/turbo-v2.5", "elevenlabs"),
    text: "Hello! Welcome to our video tutorial.",
    voiceId: "21m00Tcm4TlvDq8ikWAM", // Rachel voice
  }),
).execute();

Voice Selection

ElevenLabs provides pre-made voices and custom voice cloning.

Popular Pre-Made Voices

Voice	ID	Description
Rachel	`21m00Tcm4TlvDq8ikWAM`	Calm, professional female
Adam	`pNInz6obpgDQGcFmaJgB`	Deep, authoritative male
Josh	`TxGEqnHWrfWFTfGW9XjX`	Young, energetic male
Bella	`EXAVITQu4vr4xnSDxMaL`	Warm, friendly female

generateAudio({
  model: audioModel("elevenlabs/turbo-v2.5", "elevenlabs"),
  text: "Your narration text here.",
  voiceId: "pNInz6obpgDQGcFmaJgB", // Adam
});

Options

Option	Type	Default	Description
`text`	`string`	required	Text to convert (max 40,000 chars)
`voiceId`	`string`	required	ElevenLabs voice ID
`modelId`	`string`	`eleven_turbo_v2_5`	ElevenLabs model variant
`stability`	`number`	`0.5`	Voice consistency (0-1)
`similarityBoost`	`number`	`0.75`	Voice similarity (0-1)
`style`	`number`	`0`	Style exaggeration (0-1)
`useSpeakerBoost`	`boolean`	`true`	Enhanced audio quality
`outputFormat`	`string`	`mp3_44100_128`	Audio format
`languageCode`	`string`	auto	Language code

Voice Settings

Stability

Controls consistency vs expressiveness:

generateAudio({
  model: audioModel("elevenlabs/turbo-v2.5", "elevenlabs"),
  text: "This is a test.",
  voiceId: "21m00Tcm4TlvDq8ikWAM",
  stability: 0.7, // More consistent, less expressive
});

Low (0.0-0.3): More expressive, variable delivery
Medium (0.4-0.6): Balanced
High (0.7-1.0): Very consistent, robotic at extremes

Similarity Boost

How closely to match the original voice:

generateAudio({
  model: audioModel("elevenlabs/turbo-v2.5", "elevenlabs"),
  text: "Match the voice closely.",
  voiceId: "21m00Tcm4TlvDq8ikWAM",
  similarityBoost: 0.9, // Very close to original
});

Style

Exaggerate speaking style (v2 models):

generateAudio({
  model: audioModel("elevenlabs/turbo-v2.5", "elevenlabs"),
  text: "An exciting announcement!",
  voiceId: "21m00Tcm4TlvDq8ikWAM",
  style: 0.5, // More expressive delivery
});

Multi-Language

Supports 32 languages:

// Spanish
generateAudio({
  model: audioModel("elevenlabs/turbo-v2.5", "elevenlabs"),
  text: "Hola, bienvenidos a nuestro canal.",
  voiceId: "21m00Tcm4TlvDq8ikWAM",
  languageCode: "es",
});

// French
generateAudio({
  model: audioModel("elevenlabs/turbo-v2.5", "elevenlabs"),
  text: "Bonjour, bienvenue sur notre chaîne.",
  voiceId: "21m00Tcm4TlvDq8ikWAM",
  languageCode: "fr",
});

Output Formats

Format	Description
`mp3_44100_128`	MP3 128kbps (default)
`mp3_44100_192`	MP3 192kbps (higher quality)
`pcm_16000`	PCM 16kHz
`pcm_22050`	PCM 22kHz
`pcm_24000`	PCM 24kHz
`pcm_44100`	PCM 44.1kHz

Context for Better Pronunciation

Provide surrounding text for better prosody:

generateAudio({
  model: audioModel("elevenlabs/turbo-v2.5", "elevenlabs"),
  text: "This is the middle part.",
  voiceId: "21m00Tcm4TlvDq8ikWAM",
  previousText: "Here's the introduction.",
  nextText: "And here's the conclusion.",
});

With Video Generation

Create narrated videos:

import {
  compose,
  generateVideo,
  generateAudio,
  videoModel,
  audioModel,
} from "@synthome/sdk";

const execution = await compose(
  generateVideo({
    model: videoModel("veed/fabric-1.0", "fal"),
    image: "https://example.com/speaker.jpg",
    audio: generateAudio({
      model: audioModel("elevenlabs/turbo-v2.5", "elevenlabs"),
      text: "Welcome to today's tutorial on AI video generation.",
      voiceId: "21m00Tcm4TlvDq8ikWAM",
    }),
  }),
).execute();

Model Variants

Model ID	Speed	Quality	Languages
`eleven_turbo_v2_5`	Fastest	High	32
`eleven_multilingual_v2`	Fast	Highest	29
`eleven_flash_v2_5`	Very fast	Good	32

generateAudio({
  model: audioModel("elevenlabs/turbo-v2.5", "elevenlabs"),
  text: "Using the multilingual model for highest quality.",
  voiceId: "21m00Tcm4TlvDq8ikWAM",
  modelId: "eleven_multilingual_v2",
});

ElevenLabs Turbo

On this page