Bring a portrait to life with your audio. The model automatically syncs lips and expressions to the speech — a talking avatar up to 5 minutes long.
▶ Example: talking avatar with sound
A photo of a face or character (JPEG/PNG). This is what comes to life and speaks.
Speech, voiceover or a voice recording up to 5 minutes. You can record it with a microphone.
The model syncs lips and expressions to the audio and creates a talking avatar.
No recording yet? Turn your text into speech with AI — in your own voice or a voice from a description. More about text to speech →
Price = rate × audio duration
Maximum facial detail
It's a model that brings a portrait to life using your uploaded audio — creating a talking avatar with automatic lip-sync.
MP3, WAV, AAC, OGG or MP4 with speech/voiceover, up to 5 minutes long. The video length equals the audio length.
The prompt is optional. It describes the avatar's behavior and emotions — expressions, gaze, nods. The words come from the audio, not the prompt.
Per-second pricing: the per-second rate for 720p or 1080p multiplied by the audio duration.
The model syncs lips to audio in any language. For the behavior prompt, English works best.