Presenter and spokesperson clips
Pair a recording of your script with a portrait to create a short presenter video for a campaign or product update.
Talking Video
AI Talking Video Generator from a Photo
Upload a clear face photo and spoken audio, then generate a short talking-head video with synchronized mouth movement.
Talking Video: supported inputs and settings
Inputs and settings
- Portrait image (required): Upload the person who should speak.
- Mask image (optional) (optional): Optional PNG, JPEG, or WebP mask.
- Speech audio (required): The audio duration sets the generated video length.
- Resolution: 480p, 720p
- Seed: -1–2147483647
- Input
- A front-facing portrait, spoken audio, and an optional mask image.
- Output
- A 480p or 720p talking-head video whose duration follows the audio.
How to use Talking Video
01
Upload a clear portrait
Use a front-facing JPG or PNG with one visible person and an unobstructed mouth. Add a mask only when the editable region needs tighter control.
02
Add your audio
Choose clean speech in a supported format, then select 480p or 720p. The audio duration sets the output length.
03
Generate and check lip sync
Submit after reviewing the estimate, then inspect mouth timing, facial edges, eye movement, and audio quality.
Talking Video use cases
Educational explainers
Create concise lessons, announcements, or onboarding messages from a portrait and narration.
Avatar messages
Give a consistent virtual character a voice for community, support, or creative storytelling.
Input requirements and limitations
Prepare your input
Face the camera
Use one well-lit face looking toward the camera. Keep the mouth, chin, and cheeks visible.
Record clean speech
Use one speaker, clear pronunciation, and minimal music or ambient noise behind the voice.
Know the limits
Extreme angles reduce sync
Side profiles, covered mouths, and large head turns can make lip movement less accurate.
Audio rights still matter
Use a voice recording and portrait you have permission to process and publish.
Talking Video vs alternatives
| Option | Best for | Key difference |
|---|---|---|
| AI Talking Video | Animating a portrait with speech | Uses a face image and audio to create a lip-synced talking clip. |
| Video Translator | Translating an existing video | Starts from a complete video and adapts its language and voice. |
| Video Face Swap | Changing a face in an existing video | Replaces a target face rather than generating speech from a still image. |
Frequently asked questions
Use a front-facing, well-lit head-and-shoulders photo with one person and a clearly visible mouth.
Yes, if you have the rights to the avatar image and the result does not impersonate someone without consent.
Use clean speech from one speaker in a supported format. Avoid heavy background music and overlapping voices.
Only publish material you have rights and consent to use, including the portrait, voice, script, and any third-party brand elements.