About
TwinSpeak lets you clone any voice in seconds. Upload a sample, write a script, and get studio-quality speech output — powered by state-of-the-art voice synthesis.
Frequently asked questions
What is TwinSpeak?
TwinSpeak is a voice cloning tool. You upload a short audio sample and a script, and TwinSpeak generates a realistic voice clone reading that script.
How long does a voice sample need to be?
A few seconds is enough to get a basic clone, but 15–30 seconds of clear speech will produce noticeably better results. Background noise reduces quality, so a quiet recording environment helps.
Can I clone someone else's voice?
Only with their explicit consent. You must own or have the legal right to use any voice sample you upload. Cloning voices without consent violates our Terms of Service and may be illegal.
What audio formats are supported?
We accept most common audio formats including MP3, WAV, M4A, OGG, and WebM. The file size limit is 50 MB.
How many generations do I get?
Anonymous users get 1 free generation. Signed-in users on the free plan get 10 per month. Pro ($9/mo) includes 150, and Unlimited ($29/mo) includes ~1,000.
Are my voice samples stored permanently?
Audio files are stored on Cloudflare R2. You can request deletion of your data at any time by emailing legal@twinspeak.ai.
When will Pro and Unlimited plans be available?
Paid plans are coming soon. Sign in to stay updated — we'll notify users when subscriptions go live.