ComfyUI Node
Add Timed Video Reference for MiniMax H3 (Experimental)
Adds a Qwen-only video reference starting at a target time. Core VIDEO is streamed by real presentation timestamps; IMAGE batches remain available as a constant-FPS fallback.
Add Timed Video Reference for MiniMax H3 (Experimental)
- clip
- video_frames
- source_video
- CLIP
◄prompt_tagmotion►
◄target_start_seconds0.000►
◄source_fps24.000►
◄video_size512►
◄analysis_fps2.0►
Categorymodel/conditioning/minimax
Inputs (8)
| Name | Type | Default | Description |
|---|---|---|---|
| clip | CLIP | — | |
| prompt_tag | STRING | motion | Use this name with a leading # in the Ref2VA prompt, for example #motion. |
| target_start_seconds | FLOAT | 0.0000–9999 | Where this reference clip begins on the generated video's timeline. Sub-second values are preserved. |
| source_fps | FLOAT | 24.0000.01–240 | Used only by the IMAGE-batch fallback. It must match the loaded frame rate; core VIDEO reads timing automatically. |
| video_size | COMBO | 512 | Approximate pixel-area edge used for Qwen vision encoding. 64 is the technical floor but loses detail. |
| analysis_fps | FLOAT | 2.00.5–240 | Frames per second presented to Qwen. 2 matches core H3; higher rates cost more tokens and memory. Do not exceed the source rate. |
| video_framesopt | IMAGE | Legacy/VHS fallback: a decoded IMAGE batch with constant frame spacing. Connect either this or source_video, not both. | |
| source_videoopt | VIDEO | Preferred: connect core Load Video. The source stays path-backed and only selected frames become tensors; real presentation timestamps are preserved. |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| CLIP | CLIP | — |