Elevenlabs Eleven V4 API Documentation

Elevenlabs Eleven V4 API Documentation

Playground

Eleven V4 Text-to-Speech converts text into expressive multilingual speech with configurable voice selection, stability, and similarity controls for narration, dialogue, localization, virtual assistants, and voice production workflows. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.

Features

Eleven V4 converts text into expressive, natural-sounding speech for narration, voiceovers, and character performances. Choose a preset or custom voice, guide delivery with inline audio tags, and adjust voice stability and similarity to shape the final performance.

It is suited to high-quality spoken content where expressive delivery, multilingual speech, and voice consistency matter more than the lower-latency focus of the Turbo variant.

Why Choose This?

• Expressive text-to-speech
  Generate natural-sounding speech for narration, dialogue, and performance-driven content.

• Audio-tag control
  Guide delivery with inline cues such as [whispering] and [laughing].

• Preset and custom voices
  Choose a suggested voice or provide a custom voice_id.

• Multilingual speech
  Generate spoken content across multiple languages.

• Voice controls
  Adjust stability and similarity to tune delivery consistency and resemblance to the selected voice.

• MP3 output
  Receive the generated speech as a downloadable audio file.

Parameters

How to Use

1. Enter the text — Provide the script you want spoken.
2. Choose a voice — Select a suggested voice or provide a custom voice ID.
3. Add expressive cues optional — Use supported audio tags such as [whispering] or [laughing] where needed.
4. Adjust voice controls optional — Set stability and similarity to tune the performance.
5. Submit — Generate the speech and retrieve the MP3 audio file.

Pricing

Pricing is $0.08 per 1,000 input characters, prorated according to the actual length of text.

Billable characters include the complete text field, including spaces, punctuation, and inline audio tags.

Requests shorter than 1,000 characters are billed proportionally rather than as a full 1,000-character block.

The calculated charge is rounded up to the nearest $0.000001.

voice_id, stability, and similarity do not add separate charges.

Best Use Cases

• Narration — Generate polished speech for stories, explainers, tutorials, and educational content.
• Voiceovers — Create spoken audio for videos, advertisements, presentations, and social content.
• Character performances — Use expressive tags and voice controls for dialogue and performance-focused speech.
• Multilingual content — Generate spoken content for international and localized workflows.
• Creator content — Produce voice tracks for podcasts, videos, shorts, and other media.
• High-quality speech generation — Use the standard Eleven V4 model when expressive output is the priority.

Pro Tips

• Write scripts in a natural, conversational style for more fluid speech.
• Use expressive audio tags only where the delivery should clearly change.
• Use stability to adjust how consistent the performance remains across the script.
• Adjust similarity when you want to tune how closely the result follows the selected voice.
• Use punctuation intentionally because it can affect pauses and phrasing.
• Split longer scripts into separate requests when different sections need different voices or delivery styles.

Notes

• text supports 1–10000 characters.
• Spaces, punctuation, and audio tags count toward pricing.
• Requests are billed proportionally by character count.
• Voice availability depends on access to the selected voice.
• Audio tags guide delivery but do not guarantee an exact performance.
• SSML, separate style controls, and speed controls are not exposed by this endpoint.
• The endpoint returns an audio file rather than a live audio stream.

Related Models

• Eleven V4 Turbo — Generate expressive speech with lower-latency text-to-speech generation.
• Eleven V4 — Generate expressive text-to-speech with the standard Eleven V4 model.

Authentication

For authentication details, please refer to the Authentication Guide.

API Endpoints

Submit Task & Query Result

set -euo pipefail

export WAVESPEED_API_KEY="your-api-key"

REQUEST_BODY=$(cat <<'JSON'
{
  "text": "[warmly] Welcome to a new day. Take a deep breath, find your rhythm, and make a little room for something wonderful.",
  "voice_id": "Alice",
  "stability": 0.5,
  "similarity": 0.75
}
JSON
)

# 1. Submit the prediction.
SUBMIT_RESPONSE=$(curl --silent --show-error --fail-with-body \
  -X POST "https://api.wavespeed.ai/api/v3/elevenlabs/eleven-v4" \
  -H "Authorization: Bearer ${WAVESPEED_API_KEY}" \
  -H "Content-Type: application/json" \
  -d "${REQUEST_BODY}")

TASK=$(printf '%s' "${SUBMIT_RESPONSE}" | jq 'if type == "object" and has("data") then .data else . end')
PREDICTION_ID=$(printf '%s' "${TASK}" | jq -r '.id // empty')
if [ -z "${PREDICTION_ID}" ]; then
  printf 'Submission response did not contain a prediction id
' >&2
  exit 1
fi
RESULT_URL="https://api.wavespeed.ai/api/v3/predictions/${PREDICTION_ID}/result"

# 2. Poll until the prediction finishes.
while true; do
  RESPONSE=$(curl --silent --show-error --fail-with-body \
    "${RESULT_URL}" \
    -H "Authorization: Bearer ${WAVESPEED_API_KEY}")
  RESULT=$(printf '%s' "${RESPONSE}" | jq 'if type == "object" and has("data") then .data else . end')
  STATUS=$(printf '%s' "${RESULT}" | jq -r '.status // empty')

  case "${STATUS}" in
    completed) printf '%s\n' "${RESULT}" | jq '.outputs'; break ;;
    failed|cancelled|timeout|deleted) printf '%s\n' "${RESULT}" | jq . >&2; exit 1 ;;
    *) sleep 2 ;;
  esac
done

Parameters

Task Submission Parameters

Request Parameters

Response Parameters

Result Request Parameters

Result Response Parameters