GPT Image 2.5 is live — OpenAI's newest image model, targeted edits that leave the rest of the frame alone

Voice Cleaner API

Clean noisy speech before publishing or transcription. The voice cleaner API returns clean voice output with optional noise and echo controls.

Highlights
Noise & reverb removal · 3 cancelling levels
Best for
Podcasts, interviews, field recordings
Input
Audio file URL
Output
Cleaned audio
Input
0/1 file

MP3 / WAV / AAC / OGG

0

0, 1, or 2

Default auto

Optional

mp3, wav, flac, aac, or ogg

Result

Pricing

Credit-based — 1 credit = $0.001 USD. Pay only for completed generations.

First test cost
$0.2200 credits

One sample at the cheapest tier (1 started minute).

Testing budget guide
Add credits
$10
≈ 50 tests
$50
≈ 250 tests
$100
≈ 500 tests
Source speech
$0.2
Price · 1 started minute

What you can build with this model

Real-world workflows and production use cases you can build and ship with this model.

podcast speech waveform cleaned for publishing

Voice cleaner API for podcasts and interviews

Use the voice cleaner API before publishing clips, podcasts, field interviews, or webinar recordings. Reduce background noise and return speech that is easier to edit or review. The voice cleaner API is meant for speech-first audio, not music stem splitting. It helps production teams clean uploads before trimming, clipping, captioning, or sending audio to another service. The voice cleaner API also gives user-generated content products a predictable cleanup step before human review.

Clean a voice
clean speech prepared for transcription

Voice cleaner API before transcription

Noisy speech hurts transcripts and review tools. Put the voice cleaner API before your transcription pipeline so the next service receives cleaner voice audio. This is especially useful when uploads come from laptops, meeting rooms, phones, or creator microphones. A cleaner file can make speaker review easier even when the final transcript service stays the same. The voice cleaner API keeps this step async, so long recordings can finish in the background.

voiceover audio cleaned for production

Voice cleaner API for user submissions

Screen recordings, support calls, and creator submissions arrive with different rooms and microphones. Run the voice cleaner API as an async prep step before storage or moderation. Your app can keep the original file and attach the cleaned output after polling finishes. That gives reviewers a cleaner version without losing the source file for audit, support disputes, quality checks, comparison, or reprocessing later. Keep both versions available when a customer asks for proof.

Why reAPI for Voice Cleaner API

Why this voice cleaner API stays speech-focused

The voice cleaner API uses the voice cleanup model and keeps the page centered on speech, not music separation. It returns clean voice output for production workflows. That focus makes the playground easier for users who just want noise reduction.

Three noise levels

Set noise_cancelling_level to 0, 1, or 2. That gives product teams a small, clear control instead of exposing a long list of tuning knobs. Start with the middle option in the UI and let advanced users raise or lower cleanup only when they hear a reason. The voice cleaner API keeps this decision small enough for support agents, creators, and reviewers to understand.

Voice cleaner API with the same task shape

Create a task, poll the task id, then download output URLs. The voice cleaner API follows the same contract as the rest of the reAPI async endpoints. Teams can reuse existing queue, retry, and webhook-side storage patterns.

Integrate the voice cleaner API in three steps

  1. step 01

    Create an API key

    Generate a reAPI key from workspace settings. Use it from your backend to submit speech cleanup jobs.

    Open
  2. step 02

    Submit a cleanup task

    POST model audio-voice-clean with a public audio_url. Add noise_cancelling_level, dereverb_enabled, and encoder_format when needed.

    Open
  3. step 03

    Poll and download clean speech

    GET /api/v1/tasks/:id until completed. Download the clean voice output from output.audio_urls or use output.tracks when labels are present.

    Open
docs/api/voice-cleaner

API reference

Drop-in code and the full parameter table.

View full reference
curl https://reapi.ai/api/v1/audio/generations \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"audio-voice-clean","audio_url":"https://cdn.example.com/podcast.wav","noise_cancelling_level":1,"encoder_format":"mp3"}'

Parameter validation failed

Rejected synchronously with the failing field named. Check enum values, ranges, and URL fields — nothing is charged.

Authentication or balance issue

401 means a missing or invalid Bearer key; 402 means the reserve exceeds your balance. Manage keys and credits in the console.

Content or material rejected

Reference material and outputs pass automated moderation. Rejected tasks fail with a clear error and a full refund.

Task failed or timed out

A task that reaches a failed state is never charged — the reserve refunds automatically. Keep the task id and retry when ready.

Frequently asked questions

Common questions about this model.

The voice cleaner API is an async speech enhancement endpoint for reducing noise in voice recordings. Send a public audio_url, choose an optional noise_cancelling_level, and poll until clean voice output is ready. Use the voice cleaner API before publishing, transcription, moderation, or review. It is designed for backend workflows where the user should see progress instead of waiting on a blocking upload screen.

Related models

Explore more models in the same category.

View all models
start building

Ready to ship?

Try it in the playground or grab an API key to integrate now.

Try voice cleaningView voice cleaner docs