What can you create with Speech to Video?
Speech to video starts with an existing voice recording rather than generating a soundtrack from text. It is useful for testing a character introduction, a narrated portrait or a short presenter concept. A clear face and clean recording give the model more useful input; review mouth movement and expression in the result.
Talking portraits
Bring a still character portrait into a speaking scene using your uploaded voice recording.
Presenter concepts
Test how a character delivers a short introduction or explanation before building a longer video.
Character voice tests
Compare different recordings with the same image to explore tone, pacing and delivery.
How to use Speech to Video
Upload a character image
Choose a clear image with a visible face and minimal obstructions around the mouth.
Add your speech recording
Upload audio with a clear voice. Use the optional prompt to describe the scene or expression and review available settings.
Generate the speaking video
Check the displayed credit estimate, sign in and submit. Preview the video to assess speech timing and facial motion.
Generation requires an account and sufficient credits. Costs vary by model and settings; review the estimate in the form and the pricing options before submitting.
Speech to Video prompt example
Use this sample as a starting point and adapt it to your own scene and inputs.
A presenter faces the camera in a softly lit studio, speaking calmly with a relaxed expression and subtle head movement. Keep the camera steady.
Use a recording with one clear speaker and little background noise. The prompt describes the visual performance; the uploaded audio supplies the speech.
Speech to Video FAQ
What do I need for speech to video?
The Wan 2.2 speech-to-video model requires a character image and an audio file. A text prompt is optional and can guide the scene.
Does the tool turn typed text into speech?
This workflow takes an uploaded audio recording. Prepare the speech audio first, then upload it with your image.
Will the lips match the recording perfectly?
The model uses audio to guide the performance, but lip movement and expression can vary. Preview the output and check timing before sharing it.
Related AI tools
Video with Audio
Create video with audio from a text prompt or an image using Veo 3. Describe both the scene and its sound, then enable sound in the generator before submitting.
Image to Video
Animate a photo, illustration or product image with Wan image to video AI. Upload your starting image and describe the subject movement and camera motion you want.
Wan2.2 Animate
Animate a character image using a driving video with Wan 2.2 Animate. Supply the character appearance and the source performance to guide a new animation.