ComfyUI Node
SolarRenderer Conditioning (Structure to Video)
SolarRenderer ref2va conditioning: the structure video is encoded on the target canvas and packed on the target's own timeline/grid (aligned observation), the first frame as a keyframe at frame 0; the prompt reaches Qwen3-VL as <Video 1>/<Video 2>.
SolarRenderer Conditioning (Structure to Video)
- clip
- vae
- structure_video
- first_frame
- positive
- latent
◄prompt►
◄width1344►
◄height768►
◄length158►
◄fps24►
Categoryconditioning/video_models
Inputs (9)
| Name | Type | Default | Description |
|---|---|---|---|
| clip | CLIP | — | |
| vae | VAE | MiniMax H3 video VAE (e.g. minimax_h3_video_vae_fp16.safetensors) | |
| prompt | STRING | Six-part SolarRenderer prompt (skills/solarrenderer-caption). References <Video 1> (structure) / <Video 2> (first frame). | |
| structure_video | IMAGE | Structure clip frames at 24 fps, at least `length` frames; frame-aligned with the video being rendered. | |
| width | INT | 134432–16384 | — |
| height | INT | 76832–16384 | — |
| length | INT | 1585–3600 | Frame count at 24 fps, snapped up to the 17k+5 grid (158 = 6.6 s, SolarRenderer's training length). |
| fps | INT | 245–120 | Frames per second for the video being rendered. |
| first_frameopt | IMAGE | Optional first frame in the target look (<Video 2> of the prompt). |
Outputs (2)
| Name | Type | Description |
|---|---|---|
| positive | CONDITIONING | — |
| latent | LATENT | — |