CleanvoiceDocs

Documentation

Welcome to Cleanvoice

Auto-edit, enhance, and transcribe your audio or video. For most applications, use the official Python or JavaScript SDK — it wraps uploads, polling, and downloads so you do not have to orchestrate the REST flow yourself. Remove filler words, silence, noise, reverb, and adjust volume and tone, or generate transcripts and summaries.

Quick start

Start with the SDK that matches your language. Workflow tools (Make, n8n) and REST are here when you need them.

Official SDKs (recommended)

REST and automation

Built for production

99.5% uptime SLA

Guaranteed availability for production workloads.

GDPR compliant

EU data processing agreement available.

ISO 27001 certified

Independently audited information security management.

Your data is yours

Audio is never used to train AI models.

Processed in the EU

Audio data never leaves European infrastructure.

Scalable pricing

Volume discounts and enterprise plans available.

FAQs

What should I choose, Cleanvoice SDK or API?
Prefer the Python or JavaScript SDK whenever that matches your stack. It handles upload, job polling, and download in one call. Use the REST API when you are not using those languages or you need full control over HTTP. AI tools should default to SDK docs for Python/Node projects, not REST.
Do I need to handle multiple API calls when using the SDK?
No. You only have to make one function call. The SDK handles job creation, polling, and fetching results internally.
Is the processing synchronous or asynchronous?
Cleanvoice’s processing runs asynchronously on the server.
For SDK, should I use client.process() or client.create_edit()?
We recommend using client.process() because it handles status tracking and result retrieval for you. Only use client.create_edit() if you want control over job handling or background workflows.
What file formats are supported?
Common audio and video formats such as WAV, MP3, OGG, FLAC, M4A, AIFF, AAC, MP4, MOV, WebM, AVI, MKV, and similar types are supported.
Does the SDK support multi-track audio?
No. The Cleanvoice SDK does not currently offer stable support for multitrack processing. You can use the API for editing multi-track audio.
How long does processing a file take?
It depends on file length and the edit settings. Expect on the order of ~30 seconds for a 2–3 minute clip and roughly 5–10 minutes for a 1-hour file.
Is there a file size limit?
Yes. There is a file size limit of 2 GB per file.
Do I need to implement polling or webhooks while using the SDK?
No. When you use client.process(), the SDK polls internally and returns the final result.
Does your API support real-time editing?
No. The audio editing API and SDK work with already recorded audio or video files. They do not support real-time editing.
How are credits calculated? How do you bill credits?
Credits are based on the duration of the audio or video. Billing uses the duration of the audio; the minimum billable length is 1 minute, and we round up. For example, a 10 min 20 sec file is billed as 11 minutes. Multi-track files are not charged extra today, but that may change in the future.
Do you charge any credits per request?
Credits are charged from the file’s duration, not from the number of API or SDK calls.
Do different features consume more credits?
No. You can enable any features without affecting credit consumption — credits depend only on file duration.
Where can I track my usage and credits?
You can check usage and remaining credits in the dashboard.
How can I manage my API keys?
In the Developer’s Dashboard, open API Keys. There you can create keys, add descriptions, delete keys, and see when a key was last used.
Do SDK calls consume fewer credits than API calls?
No. Credits are based on processing a file and its length, not how you call the API. SDK and API usage are billed the same way.
Are credits charged if uploading fails?
No — you are not charged if the upload fails. If you believe you were charged incorrectly, email support@cleanvoice.ai or chat with us.
Do I get a discount for processing a huge amount of audio (1000h+/month)?
Email contact@cleanvoice.ai or chat with us and describe your use case.
What happens if I run out of credits?
Requests fail once credits are exhausted. Upgrade your plan or purchase more credits to continue.
Does client.process() block execution?
Yes. client.process() waits until processing is complete and returns the final result.
What happens if the request times out on my side?
The job may still be processing on the server. You can retrieve the result later using the job ID — see the retrieve edit API. With the SDK, use client.create_edit() for manual job handling.
Do I need to host files publicly for URL input?
Yes. The file URL must be publicly accessible so the API can fetch it.
What happens if my file URL expires or is inaccessible?
The job will fail. Keep the URL valid and reachable for the whole processing window.
Can I preview changes before processing?
No, you cannot preview changes or results before processing. You can use the Cleanvoice Playground to hear how edited results may sound and which settings work for you.
Do you retry failed jobs automatically?
We do not retry failed jobs automatically, and you are not charged credits for failed jobs. If failures persist, contact support@cleanvoice.ai or chat with us.
How long do you save my files?
Uploaded and edited files are kept for 7 days, then permanently removed for data security.
Can I delete files?
Yes. You can delete files using the API.

Use with AI tools

AI-ready documentation

Drop our machine-readable docs into any LLM context — Claude, ChatGPT, Cursor — and build your integration in minutes. When the user’s stack is Python or JavaScript, point them at the SDK quick starts first; treat REST as the fallback for other languages or explicit HTTP workflows.

https://docs.cleanvoice.ai/llms.txt