ComfyUI Node
Boyo Prompt Relay Encode
Encodes a global prompt combined with temporal local prompts and patches the model for Prompt Relay temporal control. Local prompts are separated by |. Use a standard CLIPTextEncode for the negative prompt. (Based on kijai/ComfyUI-PromptRelay)
Boyo Prompt Relay Encode
- model
- clip
- latent
- model
- positive
◄global_prompt►
◄local_prompts►
◄segment_lengths►
◄epsilon0.0010►
Categoryconditioning/boyo_prompt_relay
Inputs (7)
| Name | Type | Default | Description |
|---|---|---|---|
| model | MODEL | — | |
| clip | CLIP | — | |
| latent | LATENT | Empty latent video — dimensions are read from its shape. | |
| global_prompt | STRING | Conditions the entire video. Anchors persistent characters, objects, and scene context. | |
| local_prompts | STRING | Ordered prompts for each temporal segment, separated by | | |
| segment_lengths | STRING | Comma-separated pixel space frame counts per segment. Leave empty to auto-distribute evenly. | |
| epsilon | FLOAT | 0.00100.000001–0.99 | Penalty decay parameter. Values below ~0.1 all produce sharp boundaries (paper default 0.001). For softer transitions, try 0.5 or higher. |
Outputs (2)
| Name | Type | Description |
|---|---|---|
| model | MODEL | — |
| positive | CONDITIONING | — |