Wan Prompt Loop Info

Chained Wan generation is the standard way to get past the 5-second clip ceiling, and it's a math trap: every prompt in your loop becomes a generation, every generation has a frame count, and the total runtime sneaks up on you. IAMCCS_WanPromptLoopInfo is the pre-flight check - paste your prompt bank, tell it your frame budget and FPS, and it tells you how many generations, frames, and seconds the whole loop will cost.

What it computes

text is your prompt bank - multiple prompts, split according to separator_mode: auto (it figures out the format), indexed, blank_line, or line. Then:

• first_visible_frames (default 81) - frames for the bootstrap generation.
• continuation_visible_frames (default 81) - frames for each follow-on generation.
• fps (default 16 - the Wan-native default) - for the duration math.
• bootstrap_outside_loop (default true) - if true, the first prompt is the bootstrap and the rest are continuations: continuation_generations = prompt_count - 1 and total_frames = first + (count-1) * continuation. If false, every prompt is a continuation.

Outputs: prompt_count, total_generations, continuation_generations, total_frames, total_duration_s, and a one-line report that summarizes everything - handy to drop into a text preview.

The honest use

This is a planning node for the pack's Wan loop workflows (it lives in the same file family as IAMCCS_WanLongPlanner). The value is knowing the scale before you queue: 10 prompts at 81 frames and 16fps is roughly 50 seconds of video and, depending on your GPU and model, a long wall-clock time. Seeing total_duration_s and total_frames up front is how you decide to cut the bank from 10 prompts to 4. It does exactly the arithmetic - no sampling, no magic - so the only real risk is feeding it a separator format it doesn't parse the way you expect, which is what the report's prompt_count is for: glance at it and you'll know instantly if the split went wrong.

Install

Part of IAMCCS-nodes:

• ComfyUI Manager → search "IAMCCS" → install → restart.
• Or:

cd ComfyUI/custom_nodes
git clone https://github.com/IAMCCS/IAMCCS-nodes.git

Restart ComfyUI. Requirements: ComfyUI ≥ 0.3.0, Python ≥ 3.12, Torch ≥ 2.8. Pure math, no models.