Sync React 1: the mouth pass after the picture exists (84cr)
Sync React 1 wants a plate and a track. It is not how you invent the scene.
Sync React 1 wants a plate and a track. It is not how you invent the scene. The catalog description: “Emotional lipsync with facial expressions and head movement. Supports emotions (happy, angry, sad, neutral, disgusted, surprised), model modes (lips, face, head), and temperature control.” Category is video-to-lipsync. Content type is lipsync. Audio is on. Credits: 84. No duration list. No resolution token. Featured in the snapshot.
Sync React 1 is a mouth pass. The room, the wardrobe, the cut, the voice file — those have to exist. If you do not have a plate, generate one on a video or image-to-video row. If you do not have a track, write one on a text-to-speech row. Then open Sync. Asking React 1 to “make a person in a kitchen say the script” is an 84-credit way to not invent a kitchen.
Emotions and modes are after the picture exists
Emotions listed: happy, angry, sad, neutral, disgusted, surprised. That is a performance control on a face you already have, not a scene prompt. Modes: lips, face, head. Lips is the small pass — jaw and mouth, keep the performance in the plate. Face adds expression. Head adds head movement. Temperature is how wild the pass is allowed to get. High temperature on a locked spokesperson is how you get a different person with the same haircut.
Video-to-lipsync means the input shape is a clip (a performance plate), not a still-only talking generate. If you only have a still, you are in a different lipsync flavour — not this row’s category. The AI lipsync tool is the door that routes the job. The editing task is lipsync video. The ranked list is best lipsync model.
The longer “mouth is the product” argument is lipsync when the mouth is the product. This page is the 84-credit React 1 endpoint: emotions, lips/face/head, temperature, Sync.
84 credits is a mouth pass, not a film
The catalog sticker is 84 credits. The price matrix is per second (17 in the snapshot) with a floor that lands at 84. There is no 5s/10s menu on the record. Shorten the plate and the track to the line you need. An 84-credit pass on a 30-second ramble is a long emotional journey you will cut anyway.
Native audio is on in the sense that this job is audio-driven. The track is the clock. The plate must have a mouth that can open. Laying React 1 on a side-of-head shot is how you spend 84 credits on a cheek. Frame the face. Light the mouth. Then pick an emotion that matches the line — angry on a warranty disclaimer is a different commercial.
Do not invent a resolution. None is listed. You get the plate’s pixels back with a new mouth. If you need 4K, upscale the plate before or after, as a different row. React 1 is not an upscaler and not a 1080p generate.
Invent the scene elsewhere. Buy the mouth here.
Pipeline:
- Lock the picture (stills row, then image-to-video, or a talking generate you already like except the line).
- Lock the track (MiniMax Speech or another TTS, or a recorded VO).
- Run Sync React 1 with an emotion, a mode, a temperature you can defend.
- Caption and ship.
Skip step 1 and you are in a video generate. Skip step 2 and you are hoping React 1 writes a script. It will not. It syncs. 84 credits is the reason to respect that order.
Lipsync as a category is full of cheaper rows. React 1 is the emotional, head-moving, temperature-controlled pass. Use it when the face has to react, not when you needed a wav under B-roll. B-roll plus wav is voiceover. This is a mouth.
FAQ
Can Sync React 1 generate the person and the room?
No. Description is emotional lipsync on a plate with a track. Invent the scene on a video row. Come back for the mouth.
What emotions and modes are listed?
Emotions: happy, angry, sad, neutral, disgusted, surprised. Modes: lips, face, head. Plus temperature control. Pick one emotion per pass. Mixing “happy angry” in the prompt is not a listed control.
Why 84 credits?
That is the catalog sticker (featured, premium in the snapshot). The meter is per second. Shorten the line. Do not use React 1 as cheap TTS — TTS is a 2-credit voice file on a different row.
Does it need an image?
requires_image is false on this record; the category is video-to-lipsync. Plan on a plate with a visible mouth and a track. A still-only talking job is a different lipsync row.