MiniMax H3 Digital Humans in ComfyUI: One Presenter, Two Characters, One Audio Lock

Description and use

This workflow in ComfyUI enables creation of single‑ or two‑character digital human videos synchronized to audio, using MiniMax H3 and VoxCPM models.

Problem

Generate coherent digital human videos with one or two characters exchanging lines while keeping the picture stable and lip‑syncing to a single audio track.

Outcome

The provided ComfyUI graphs enable creation of single‑character and two‑character digital human videos with audio locked to the generated performance, without requiring inference during the article.

Workflow

• Load character image(s)
• Load VoxCPM reference audio
• Crop audio to desired segment(s)
• Generate spoken line with VoxCPM
• Lock audio with VRGDG_MiniMaxH3AudioDrive
• Generate video with MiniMax H3
• Create and save video output

Evidence

Two ComfyUI graphs ship with this release, both on the MiniMax H3 ref2va checkpoint with VoxCPM generating the speech and the generated (or concatenated) audio locked as the driving track.

Source

The node that shapes the whole thing is VRGDG_MiniMaxH3AudioDrive. In both saved files it is configured to lock the original audio and let H3 generate picture only.

Source