TOOL
AI lip sync: a talking video from a photo and your voice
One portrait plus your voice or a voiceover becomes a video where the person speaks or sings your words.
Exact price before Generate$10 trial · plans from $25Credits don't expire
Add a portrait and an audio file; Riftway Studio recognises the words with timing and generates a clip where the lips follow them. Your audio is kept as it is. A 5-second 720p talking clip costs 160 credits ($1.60); the price is shown before you generate.
Updated 2026-11-18How it works
- 1Add a portraitA face generated here from text works best with video models.
- 2Add the voiceYour recording, a voiceover or a song — MP3 or WAV.
- 3GenerateWords, pauses and language are detected and passed to the model.
Made in Riftway Studio
Real results from the studio. Tap to play.
What you get
Prices
1 credit = $0.01. Real prices charged today; the studio always shows the exact price before you generate.
| Model | 480p | 720p | 1080p |
|---|---|---|---|
| Seedance 2.5 | 75 cr $0.75 | 160 cr $1.60 | 412 cr $4.12 |
| Kling 4.0 | — | 145 cr $1.45 | 215 cr $2.15 |
| Wan 3.0 | 35 cr $0.35 | 70 cr $0.70 | 135 cr $1.35 |
| Seedance 2.0 | 51 cr $0.51 | 111 cr $1.11 | — |
| Seedance 2.0 fast | 41 cr $0.41 | 87 cr $0.87 | — |
| Seedance 2.0 mini | 27 cr $0.27 | 55 cr $0.55 | — |
Questions and answers
Which photos work?
Clear portraits. Video models decline photos that look like real people; faces generated here from text work reliably.
Which languages are supported?
English, Ukrainian, Spanish, German and many more — the language is detected automatically or you choose it.
Can it sing?
Yes — use a song as the audio and the singer follows the lyrics.
How long can it be?
Up to 30 seconds per clip; longer videos are joined from several clips.
Related
Try it with your own idea
Sign up, see the exact price, generate. Start with a $10 trial.
Make a talking video →