ComfyUI Node
SCAIL-2 Infinity (auto window)
All-in-one SCAIL-2 long-video node. Internally loops 81-frame chunks (with a 5-frame overlap) over the driving pose video, stitching them into one continuous video. Replaces the WanSCAILToVideo + KSampler + VAEDecode + manual chunk-wiring graph.
SCAIL-2 Infinity (auto window)
- positive
- negative
- model
- vae
- pose_video
- pose_video_mask
- reference_image
- reference_image_mask
- clip_vision_output
- images
- latent
- total_frames
◄width512►
◄height896►
◄seed0►
◄steps20►
◄cfg6.0►
◄sampler_name▾►
◄scheduler▾►
◄denoise1.00►
◄window_length81►
◄previous_frame_count5►
◄max_frames0►
◄decode_tiledfalse►
◄vary_seed_per_windowfalse►
◄pose_strength1.00►
◄pose_start0.00►
◄pose_end1.00►
◄replacement_modefalse►
Categoryvideo/scail2
Inputs (26)
| Name | Type | Default | Description |
|---|---|---|---|
| positive | CONDITIONING | — | |
| negative | CONDITIONING | — | |
| model | MODEL | — | |
| vae | VAE | — | |
| width | INT | 51232–16384 | — |
| height | INT | 89632–16384 | — |
| seed | INT | 00–18446744073709550000 | — |
| steps | INT | 201–10000 | — |
| cfg | FLOAT | 6.00–100 | — |
| sampler_name | COMBO | 44 options: euler, euler_cfg_pp, euler_ancestral, euler_ancestral_cfg_pp, heun, heunpp2, +38 | |
| scheduler | COMBO | 9 options: simple, sgm_uniform, karras, exponential, ddim_uniform, beta, +3 | |
| denoise | FLOAT | 1.000–1 | — |
| window_length | INT | 815–16384 | Frames per chunk. SCAIL-2 was trained at 81; changing this is not recommended. |
| previous_frame_count | INT | 51–16384 | Overlap frames anchored from the previous chunk. SCAIL-2 trained at 5. |
| max_frames | INT | 00–16384 | Hard cap on total output frames. 0 = run until the driving pose video is exhausted. |
| decode_tiled | BOOLEAN | false | Use tiled VAE decode to bound decode VRAM at high resolution. |
| vary_seed_per_window | BOOLEAN | false | Add the window index to the seed each chunk. Off = same seed every chunk (anchored continuity). |
| pose_strength | FLOAT | 1.000–10 | — |
| pose_start | FLOAT | 0.000–1 | — |
| pose_end | FLOAT | 1.000–1 | — |
| pose_videoopt | IMAGE | Driving pose video. Its length determines how many frames are generated. Downscaled to half resolution internally. | |
| pose_video_maskopt | IMAGE | SCAIL-2 colored per-identity SAM3 mask video, same resolution as pose_video. | |
| replacement_modeopt | BOOLEAN | false | False = Animation Mode (black-bg mask). True = Replacement Mode (white-bg mask). |
| reference_imageopt | IMAGE | Reference character image. For multiple references composite them on a single image. | |
| reference_image_maskopt | IMAGE | SCAIL-2 colored reference mask, same resolution as reference_image. | |
| clip_vision_outputopt | CLIP_VISION_OUTPUT | CLIP vision features for conditioning. |
Outputs (3)
| Name | Type | Description |
|---|---|---|
| images | IMAGE | Full stitched video. |
| latent | LATENT | Concatenated, overlap-free latent of the full video. |
| total_frames | INT | — |