FAQ
Quickly find the answers you need
Basic Questions
What is voice cloning?
Voice cloning is an AI technology that analyzes your provided audio samples to train an AI model capable of mimicking that voice's characteristics. This model can convert any text into speech with the same voice features.
What is AI voiceover? How do I create AI voiceover online?
AI voiceover is the process of converting text into human-quality speech using AI voice synthesis. On VoiceClone, simply pick a preset voice or your cloned voice model, type your text, and generate AI voiceover online with support for 30+ languages and emotion control — ideal for videos, audiobooks, and courses.
How long should the audio sample be?
We recommend providing 10-30 seconds of audio sample. The audio should be clear, without background noise, and contain natural speech expression.
How long does it take to train a model?
Typically, model training takes 1-3 minutes. Training time depends on audio file size and system load. Please keep the page open during training to avoid interruption.
Usage Questions
What audio formats are supported?
Currently supports MP3 and WAV format audio files, with a file size limit of 50MB. We recommend using audio with 44.1kHz or higher sampling rate for best results.
What is the quality of generated speech?
Speech quality depends on multiple factors: the quality of original audio sample, choice of training model, and accuracy of reference text. Using clear audio samples and S1 model typically yields the best results.
Is there a character limit for text-to-speech?
We recommend no more than 5000 characters per conversion. If you need to convert longer text, it's better to do it in segments for better speech quality.
How to download generated audio?
After generating audio on the "Text to Speech" page, click the "Download Audio" button below the player to save as MP3 format. You can also manage all generated audio files in the "My Audios" page.
Account & Security
Why do I need to log in?
Logging in allows you to save trained voice models, manage generated audio files, and ensures your data security. We support two quick login methods: Google and GitHub.
Is my data secure?
We take user data security very seriously. All uploaded audio and generated models are associated with your account and only accessible by you. We will not use your data for other purposes.
Can I delete my models and audio?
Yes. You can delete unwanted content anytime in "My Models" and "My Audios" pages. Please note that deletion is irreversible, so operate with caution.
Technical Issues
Why did training fail?
Training failure may be caused by: 1) Incorrect or corrupted audio file format; 2) Unstable network connection; 3) Reference text doesn't match audio content. Please check these factors and try again.
What if generated speech sounds unnatural?
Try these improvements: 1) Retrain model with higher quality audio samples; 2) Use punctuation properly in text; 3) Avoid special symbols and emojis; 4) Choose S1 model instead of lightweight version.