ComfyUI Node
H3 Auto Chain + Stitch
Automatically requeues the next H3 chunk, advances the Motion Context latent slots, and stitches all clips into one MP4 when the audio is complete.
H3 Auto Chain + Stitch
- video
- chain_config
- audio
- VIDEO
◄audio_sourceoriginal audio input►
◄delete_completed_latentsfalse►
Categoryconditioning/minimax
Inputs (5)
| Name | Type | Default | Description |
|---|---|---|---|
| video | VIDEO | — | |
| chain_config | H3_CHAIN | Connect the chain_config output from H3 Auto Chain Audio. This supplies the chain ID, clip timing, and latent slots. | |
| audio_source | COMBO | original audio input | Original audio input replaces the audio in the generated clips at frame-aligned boundaries. Generated video audio keeps the audio already embedded in each generated video clip. |
| delete_completed_latents | BOOLEAN | false | Delete this chain's saved H3 latent files after the final MP4 is stitched successfully. Disable to keep them available for retry or resume. |
| audioopt | AUDIO | The complete original audio. When connected, the final MP4 uses this source at frame-aligned clip boundaries instead of reusing per-clip encoded audio. |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| VIDEO | VIDEO | — |