React 1 syncs a recorded audio track to any video, generating realistic lip movements that match the spoken words frame by frame. If you shot footage and recorded audio separately, or need to dub a clip into another language, this model closes the gap without re-shooting. The result looks natural because it reads the phonetics and timing of your audio instead of just opening and closing the mouth mechanically. The emotion layer is what sets React 1 apart. You pick the mood, happy, sad, angry, disgusted, surprised, or neutral, and the model bends the facial expression around that tone while keeping the lip sync tight. You also choose which region to edit: lips only for quick corrections, the full face for expressive results, or the whole head for natural talking head movements. A temperature slider lets you dial in how subtle or intense the performance looks. React 1 fits into any video production workflow without adding extra steps. Upload a .mp4 and a .wav, choose your settings, and get a synced clip back in minutes. It works for dubbing short films, animating social media avatars, fixing a flubbed take, or producing a talking head video from a still photo paired with a voiceover.
React 1 takes a video and a separate audio track and generates realistic lip sync with per-emotion facial control. On Picasso IA, you run it directly in the browser without installing software or writing a single line of code. A common scenario: you recorded a clean voiceover after the shoot but the on-screen mouth does not match. React 1 fixes that in one pass. It also lets you assign an emotional tone to the face, so the result does not just move the lips, it gives the subject a full performance.
Do I need programming skills or technical knowledge to use this? No, just open React 1 on Picasso IA, adjust the settings you want, and hit generate.
Is it free to try? Yes, you can run React 1 without entering payment details. Try your first sync at no cost and see the result before committing to anything.
What file formats does the model accept? The model takes .mp4 for video and .wav for audio. Convert your files to those formats before uploading for the best results.
How long does it take to get results? Most clips process in under a minute. Shorter videos with a focused edit region (lips or face) tend to return faster than full head generation on longer footage.
What happens if my audio is longer than my video? Use the sync mode setting to decide how the model handles the mismatch. Loop repeats the video, bounce plays it forward and back, cut-off stops at the video's end, silence pads with a still frame, and remap stretches the video to match the audio length.
Can I adjust how expressive the result looks? Yes. The temperature setting controls how animated the facial performance is. A lower value gives subtle, restrained movement; a higher value produces more pronounced expressions that match the chosen emotion.
Where can I use the output? The downloaded clip is a clean video file with no watermarks. Drop it into any video editor, publish it to social media, or use it inside a short film or presentation.
Everything this model can do for you
Choose from six mood options to shape the facial expression around the spoken audio.
Target just the lips, the full face, or the whole head for talking head animation.
Set how subtle or pronounced the facial performance looks with a single 0-to-1 slider.
Handle audio-video length mismatches with loop, bounce, cut-off, silence, or remap modes.
Lip movements align to spoken words at the phoneme level for a natural result.
Download clean video output ready to drop into your editing timeline or publish directly.
Upload your files, pick your settings, and run the model directly in the browser.
Fast, automated pipeline for content creators
