muapi Logomuapi Logo
muapi Logomuapi Logo

ElevenLabs TTS Turbo 2.5: AI Audio Generator

Convert text to speech with ElevenLabs TTS Turbo 2.5. Fast, natural single-voice narration with tunable stability and speed. Try free — pay per generation.

📝

Overview

About this model

ElevenLabs TTS Turbo 2.5 converts written text into natural, expressive single-voice speech optimized for speed without sacrificing audio quality. Choose from a curated set of built-in voices or supply your own ElevenLabs custom voice ID, then fine-tune stability, similarity, and speaking rate to match your project's tone.

Built for high-throughput narration workloads, Turbo 2.5 is well suited for voiceovers, IVR prompts, and app-generated audio where turnaround time matters. For multi-speaker conversations with several distinct characters, use ElevenLabs Text to Dialogue V3 instead.

1Voiceover: Fast narration for videos, ads, and social content.
2IVR/telephony: Generate prompt audio for phone systems and voice assistants.
3Accessibility: Convert articles and documents to spoken audio.
4App audio: Dynamic in-app narration, notifications, and alerts.
5E-learning: Course narration and quiz feedback audio.
6Localization: Speech in supported languages for global audiences.
💰

Pricing & Value

Cost analysis

muapiapp$0.05 per 1,000 characters

Pay-per-use pricing based on trimmed text length, no subscription required.

Fal.aiNot available

Fal.ai does not currently offer ElevenLabs TTS Turbo 2.5.

ReplicateNot available

Replicate does not currently offer ElevenLabs TTS Turbo 2.5.

* Competitor pricing is estimated based on similar model architectures and usage tiers.

⚙️

Technical Details

Configuration schema

Promptstring

Text to convert to speech. Maximum 40,000 characters; billing is based on trimmed length.

Default ValueWelcome to Muapi. Your workspace is ready.
Voice IDEnum (26 options)

ElevenLabs voice ID. Select a popular voice or paste your own custom voice ID.

Default Value21m00Tcm4TlvDq8ikWAM
Stabilitynumber

Determines voice consistency and randomness (0 to 1, default 0.5).

Default Value0.5
Similarity Boostnumber

Determines how closely the output resembles the selected voice (0 to 1, default 0.75).

Default Value0.75
Speednumber

Speech rate adjustment (0.7 to 1.2, default 1.0).

Default Value1
Language CodeEnum (7 options)

Target language for speech. Leave empty for automatic language detection.

Default Valuenull
📖

Implementation Guide

Developer documentation

How to Use ElevenLabs TTS Turbo 2.5

  1. Write Your Text

    • Set prompt to the text you want spoken (up to 40,000 characters; billing is based on trimmed length).

  2. Pick a Voice (Optional)

    • Set voice_id to one of the built-in voices or your own ElevenLabs custom voice ID. Defaults to Rachel.

  3. Tune Delivery (Optional)

    • stability (0–1, default 0.5): higher keeps the voice steady, lower adds expressive variation.
    • similarity_boost (0–1, default 0.75): how closely the output matches the selected voice.
    • speed (0.7–1.2, default 1.0): speaking rate.

  4. Set Language (Optional)

    • Use language_code (e.g. en, fr, de, ja, vi, hu, no) to hint the target language.

  5. Submit the Request

curl -X POST https://api.muapi.ai/api/v1/elevenlabs-tts-turbo-2-5 \
  -H "x-api-key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "Welcome to Muapi. Your workspace is ready.",
    "voice_id": "21m00Tcm4TlvDq8ikWAM",
    "speed": 1.0
  }'
  1. Retrieve the Audio
    • Poll /api/v1/predictions/{request_id}/result until the status is completed, then download the returned audio URL.
❓

Common Questions

Frequently asked

What does ElevenLabs TTS Turbo 2.5 do?

It converts a block of text into natural-sounding, single-voice speech using ElevenLabs' Turbo 2.5 synthesis engine, optimized for fast turnaround.

How is this different from Text to Dialogue V3?

Turbo 2.5 generates single-voice narration from plain text. Text to Dialogue V3 generates multi-speaker conversations from a structured dialogue script with a distinct voice per turn.

How do I choose a voice?

Set `voice_id` to one of the built-in voice options, or paste your own ElevenLabs custom voice ID. If omitted, the model defaults to the Rachel voice.

What does the `stability` parameter control?

Stability (0–1) controls how consistent the voice sounds. Values near 1.0 keep delivery uniform and predictable; lower values allow more natural, expressive variation.

What languages are supported?

Turbo 2.5 supports English, French, German, Japanese, Vietnamese, Hungarian, and Norwegian via the `language_code` parameter. Leave it empty for automatic language detection.

How is pricing calculated?

Cost is based on the trimmed character count of your `prompt` text, billed per 1,000 characters.