Text to speech
Preview stock voices and edit your script today. Generating new audio is not open yet.
Early access: preview 100+ stock voices and sign in with Google. New audio generation, clone creation, Voice Design matching, and purchasing are not open yet.
Explore 100+ voices you can preview, mainly in English, with Chinese options and a smaller selection in other languages. Save a draft in your browser. For a closer match to your scene, Voice Design matching is coming next.
Your script is read as written. Write in the selected voice’s language.
114 matches · Selected Lucian
Audition Lucian · English
Clone your voice
One clear MP3 or WAV recording. Aim for 1–3 minutes; the first three minutes are used.
Sign in with Google to clone your voiceDescribe the scene and tone. Voice Design finds up to three matching directory voices.
Sign in with Google to find voice matchesFinding matches does not use audio minutes. Choose a result, then generate your script.
Free · 120 characters per clip
Sign in for private saving and your monthly audio allowance.
Audio generation is not open yet. You can preview voices and keep your script in this browser.
Named copies stay on this browser, with their script and voice. They are available without signing in. Restoring only fills the editor.
Overview
Explore 100+ voices you can preview, browser drafts, and the Text to Speech, Clone, and Design tabs. Google sign-in is available; new generation, clone creation, and Voice Design matching are not open yet.
Preview stock voices and edit your script today. Generating new audio is not open yet.
Explore the Clone tab. Creating a voice from an authorized recording is not open yet.
Draft a scene, accent, tone, or style in the Design tab. New voice matching is not open yet.
When generation opens, review and download your completed audio for your own editor.
Benefits
Listen to the voice library today. Creating a voice from your recording and finding matches from a description will become available later.
Explore 100+ voices you can preview, with English as the main selection, Chinese options, and fewer other languages.
Clone creation is not open yet. HeyGen’s model creates a voice from one recording; the voice must be ready before speech.
Need more voices suited to your scene? Voice Design will find up to three matching candidates from a description. New matching is not open yet.
Model at a glance
Published HeyGen model capabilities. Clone creation and new Voice Design matching are not open yet.
Model figures from HeyGen’s Instant Clone and Voice Design documentation. Explore 100+ voices you can preview today, with English at the center of the selection, plus six recorded project examples. New Voice Design matching is not open yet.
Independently operated. Powered by HeyGen voice services; not affiliated with or endorsed by HeyGen.
Features
Explore the Text to Speech, Clone, and Design tabs. Stock previews are available; new audio generation, clone creation, and Voice Design matching are not open yet.
Preview a voice and edit your script now. New generation and finished-clip downloads are not open yet.
The Clone tab is available. Creating a voice from one clear, authorized recording is not open yet.
Write a scene description in the Design tab. New matching for accent, tone, or pacing is not open yet.
FAQ
Voices, recordings, scripts, and the audio you make. A few practical answers before you start.
Use cases
A product intro, a welcome, or one clear instruction. Start with a real example and make the words your own.
A conversational opening for a product video.
English · 6.4 sec
A softer introduction for a story or welcome message.
English · 5.8 sec
An energetic opening for a short video.
English · 6.2 sec
A measured read for a lesson or walkthrough.
English · 6.1 sec
A provided Spanish script, read as written.
Spanish · 8.4 sec
A provided Chinese script, read as written.
Chinese · 11.7 sec
Explore 100+ voices you can preview and sign in with Google. Generation and purchasing are not open yet. The planned Free allowance is 900 audio seconds per rolling month, with 120 characters per clip once generation opens.
Open the voice studio