Stream text-to-speech audio

Convert text to speech using ElevenLabs AI voices with streaming for real-time playback. Returns audio data as an MP3 stream for immediate playback with minimal latency. Perfect for legal document narration, client presentations, or accessibility features.

POST /voice/v1/speak/stream

200

"<string>"

Authorizations

Authorization · string · header · required

API key starting with sk_case_

Body

application/json

text · string · required

Text to convert to speech

Maximum string length: 5000

voice_id · string · default: EXAVITQu4vr4xnSDxMaL

ElevenLabs voice ID (defaults to Rachel for professional clarity)

model_id · enum<string> · default: eleven_multilingual_v2

TTS model to use

Available options: eleven_monolingual_v1, eleven_multilingual_v1, eleven_multilingual_v2, eleven_turbo_v2

voice_settings · object

Optional voice tuning controls for delivery, consistency, and emphasis

Show child attributes

language_code · string

Language code (e.g., 'en', 'es', 'fr')

output_format · enum<string> · default: mp3_44100_128

Audio output format

Available options: mp3_44100_128, mp3_22050_32, pcm_16000, pcm_22050, pcm_24000, pcm_44100

optimize_streaming_latency · integer

Optimize for streaming latency (0-4)

Required range: 0 <= x <= 4

seed · integer

Random seed for reproducible generation

previous_text · string

Previous text for context

next_text · string

Next text for context

apply_text_normalization · boolean · default: true

Apply text normalization

enable_logging · boolean · default: true

Enable request logging

Response

200

audio/mpeg

Audio stream successfully generated

MP3 audio stream