Step 1 · Choose a voice

All 1,200 studio voices are in the box — scroll inside it. Tap a voice to select it. Press ▶ Hear and that voice itself speaks a short line and becomes your selected voice, so what you hear is exactly what will read your script.

♂/♀ is a best guess from the voice's name — HeyGen doesn't publish it. Press ▶ Hear to be sure.

Step 1B · …or clone your OWN voice

Upload one clear recording of your voice. The studio makes a copy of it that can say anything you type in Step 2. Your recording is only used to create the voice.

• 1–3 minutes of clear speech — quiet room, no music
• MP3 or WAV file, up to about 12 MB
• Speak the whole way through, like a normal conversation

2) Pick the language spoken in your recording — the studio needs it to build your voice correctly.

No recording chosen yet.

Step 2 · Write the text — get ONE voice file

Press Generate and your voice starts playing in front of you as it is being made — one continuous voice, growing until it is done, then saved as one single file to download. The engine itself speaks up to 5,000 characters (≈ 6 minutes) at a time, so for longer scripts the studio continues seamlessly behind the scenes — longest single file here: 20,000 characters ≈ 25 minutes. Keep this page open while it works.

Step 3 · Make the video

Download your voice above, choose the pictures for your video here to check them, then open the MOTION-X AutoVideo studio — put the voice and the pictures in there, and your finished video comes out. Everything stays on your own device.

No pictures chosen yet.