ComfyUI Node
Frame Context Fit
Compresses a long image sequence into a fixed-size context window. Key frames are repeated, and intermediate frames are selected with easing.
Frame Context Fit
- images
- images
- key_frame_data
- frame_indices_report
◄context_window80►
◄key_frame_interval80►
◄num_start_frames0►
◄key_frame_repeats5►
◄easing_modelinear►
◄include_last_frametrue►
Categoryimage/sequence
Inputs (7)
| Name | Type | Default | Description |
|---|---|---|---|
| images | IMAGE | — | |
| context_window | INT | 804–4096 | Total number of output frames (the context window size). |
| key_frame_interval | INT | 801–10000 | Pick a key frame every N input frames (e.g. 80 → frames 0, 80, 160 ...). When using iterative video generation with num_start_frames > 0, the first interval is this value and subsequent intervals are (key_frame_interval - num_start_frames). |
| num_start_frames | INT | 00–1000 | Number of overlapping start frames in iterative video generation. When > 0, key frames are placed at iteration boundaries: first at key_frame_interval, then every (key_frame_interval - num_start_frames) frames. Set to 0 for uniform spacing. |
| key_frame_repeats | INT | 51–100 | How many times each key frame is repeated in the output. |
| easing_mode | COMBO | linear | Easing curve applied when selecting intermediate frames between key frames. |
| include_last_frame | BOOLEAN | true | If True, the very last input frame is always included as a final key frame (with repeats), even if it doesn't land exactly on an interval boundary. |
Outputs (3)
| Name | Type | Description |
|---|---|---|
| images | IMAGE | — |
| key_frame_data | STRING | — |
| frame_indices_report | STRING | — |