ComfyUI Node
MiniMax H3 Join AV Latent / 合并音视频潜空间
Recombine video and audio latents into an H3 AV latent. Validates that both streams describe the same frame count. / 将视频和音频潜空间重新合并为 H3 音视频潜空间，并验证两个流对应相同帧数。
MiniMax H3 Join AV Latent / 合并音视频潜空间
- video
- audio
- LATENT
◄stricttrue►
Categorymodel/latent/minimax
Inputs (3)
| Name | Type | Default | Description |
|---|---|---|---|
| video | LATENT | — | |
| audio | LATENT | — | |
| strict | BOOLEAN | true | Reject streams whose latent_t and audio_t disagree, or whose latent_t is off the 17k+5 grid. Turn off only when deliberately building an irregular pair. / 拒绝时长不一致或不符合 17k+5 网格的流；仅在有意构建非标准组合时关闭。 |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| LATENT | LATENT | — |