Frequently asked questions
Quick answers about using voicechange.io. Something missing? Ask us.
Using the voice changer
Is voicechange.io really free?
Yes. Changing your voice, text to speech and song covers are all free, and so is the desktop app, with no account and no watermark. The site is paid for by ads.
Do I need to sign up or install anything?
No. Everything runs in the browser on desktop and mobile. Your browser gets a random ID so it can show your recent results and any voice you trained; clearing your cookies resets it.
How long can my recording be?
Recordings and uploads can be up to 5 minutes and 15 MB. Song covers can be up to 8 minutes. Text to speech takes up to 1,000 characters at a time (roughly a minute of speech). For longer material, split it into parts and convert them one at a time.
Which file formats can I upload?
mp3, wav, m4a, ogg and webm. If a file is too large, export it as an mp3; a 5 minute mp3 at 192 kbps is about 7 MB.
How long does a conversion take?
Usually 10 to 60 seconds, depending on the length of the clip and how many other people are converting at the same time. Song covers take longer because the vocals are separated from the music first. The page updates on its own when your result is ready.
Can I use it in real time, for example on Discord?
No. voicechange.io converts recordings and files, not a live microphone stream. Record what you want to say, convert it, then play or share the result.
Sound quality
Why does my result sound robotic or glitchy?
The most common causes are background noise or music in the recording, a pitch setting that is too far from your natural voice, and very quiet or clipped (too loud) input. Record close to the mic in a quiet room, leave auto pitch on, and see the recording tips guide for more fixes.
What does auto pitch do?
Each voice has a measured centre pitch. Auto pitch shifts your recording toward it before conversion so a deep voice going into a high voice (or the reverse) does not sound strained. You can nudge it with the slider; 12 steps is one octave.
Why do some voices sound better than others?
Each voice is a separate model trained on different amounts and quality of audio. Voices trained on lots of clean dialogue or isolated vocals sound the most convincing. Every voice page has a sample clip so you can hear the model before using it.
Can instrument voices like drums speak text?
Instrument models turn the rhythm and pitch of your voice into that instrument, so they work best with beatboxing, humming or singing. Typed text works but sounds like speech played on the instrument.
Voices and training
Can I request a voice that is not in the list?
Yes. Use the "Can't find a voice?" box under the voice picker. Requests are reviewed regularly, and a voice is added when a good public model exists for it.
Can I make an AI version of my own voice?
Yes, with the free VoiceChange desktop app, which you can download from the desktop app page (voicechange.io/real-time-voice-changer). Voice cloning is no longer done on the website. If you trained a voice here before, it is still in the "Your voice" tab, where you can keep using it or delete it.
Can other people use the voice I trained?
No. A trained voice is tied to the browser that created it and does not appear in the public catalog.
Are these the official voices of these artists and characters?
No. They are fan-made AI models that imitate those voices. voicechange.io is not affiliated with or endorsed by any of the people, characters or rights holders.
Rights, safety and privacy
Can I post the results on YouTube or TikTok?
Parody and fan content are common uses, but the rules depend on the platform and on who owns the underlying song or character. Label AI voices as AI, do not imply a real person said or endorsed something, and check the platform's policy on synthetic media and copyrighted music before monetizing.
What is not allowed?
Using a voice to deceive people into believing a real person said something, fraud or scams, harassment, and training a model on someone else's voice without permission. See the terms of use for the full list.
What happens to the audio I upload?
It is used only to produce your result and is not sold or used to train public voices. Recordings you upload for training are used only to build your private model and are deleted as soon as training finishes. The privacy policy has the details.
My voice (or my client's) is in the catalog. Can it be removed?
Yes. Email us from the contact page with the voice name and your relationship to it, and it will be taken down.