Elevenlabs Audio Isolation API Documentation

Elevenlabs Audio Isolation API Documentation

Playground

ElevenLabs Audio Isolation isolates speech from background noise to produce clearer dialogue, interviews, voice recordings, podcasts, and other speech-focused audio workflows. Ready-to-use REST inference API, best performance, no coldstarts, affordable pricing.

Features

ElevenLabs Audio Isolation separates speech from background noise to produce cleaner, more intelligible voice recordings. Upload an audio file containing dialogue or spoken content, and the model reduces surrounding noise while preserving the speech for interviews, voiceovers, meetings, and other dialogue-focused workflows.

It is designed specifically for speech isolation rather than music stem separation.

Why Choose This?

• Speech isolation
  Extract clearer spoken audio from recordings with background noise.

• Dialogue cleanup
  Improve interviews, conversations, narration, and other speech-focused recordings.

• Simple workflow
  Only an input audio file is required.

• Long-audio support
  Process recordings up to 60 minutes.

• Large-file support
  Accept audio files up to 500 MB.

Parameters

How to Use

1. Provide an audio recording — Upload a file or supply a direct audio URL.
2. Submit — Run the speech-isolation process.
3. Retrieve the result — Preview and download the cleaned audio.

Pricing

Pricing is based on the processed input audio duration.

The rate is $0.11 per input minute, charged proportionally to duration.

Only the first 60 minutes of input audio are processed and billed.

Best Use Cases

• Interviews — Reduce environmental noise around spoken dialogue.
• Voiceovers — Clean recordings captured outside controlled studio environments.
• Meetings and presentations — Improve speech clarity in recorded conversations and talks.
• Creator content — Clean dialogue for videos, podcasts, tutorials, and social content.
• Field recordings — Reduce distracting background sound around recorded speech.
• Post-production — Prepare cleaner dialogue before editing, mixing, or further processing.

Pro Tips

• Use recordings where speech is reasonably audible above the surrounding noise.
• Cleaner source recordings generally produce stronger isolation results.
• Heavy overlap between speech and loud background sounds may limit how completely the noise can be removed.
• Use the isolated output as a cleaner starting point for voice editing, mixing, or enhancement.
• Supply a direct audio-file URL rather than a webpage containing an embedded player.

Notes

• audio is the only required parameter.
• Maximum input size is 500 MB.
• Up to the first 60 minutes of audio are processed and billed.
• This endpoint is designed to isolate speech from background noise.
• It is not intended to separate music into vocals, drums, instruments, or other stems.
• Output quality depends on the source recording and the amount of overlap between speech and background noise.

Related Models

• ElevenLabs Voice Changer — Transform the voice in an existing recording.

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'
{
  "audio": "https://interactive-examples.mdn.mozilla.net/media/cc0-audio/t-rex-roar.mp3"
}
JSON
)

# 1. Submit the prediction.
SUBMIT_RESPONSE=$(curl --silent --show-error --fail-with-body \
  -X POST "https://api.wavespeed.ai/api/v3/elevenlabs/audio-isolation" \
  -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