🚫AnimateDiff Combine [DEPRECATED, Use Video Combine (VHS) Instead!] 🎭🅐🅓

Skip this one. Its own display name says it plainly: "🚫AnimateDiff Combine [DEPRECATED, Use Video Combine (VHS) Instead!]" - a warning emoji and everything, straight from Kosinkadink. This node used to be how you turned a batch of generated frames into a GIF or video file, back before ComfyUI-VideoHelperSuite existed. It still technically works if you drop it into a graph, but there's no reason to reach for it on a new workflow.

The README itself recommends ComfyUI-VideoHelperSuite (by AustinMroz and Kosinkadink jointly) "for loading videos, combining images into videos, and doing various image/latent operations like appending, splitting, duplicating, selecting, or counting" - its Video Combine node is the direct, actively-maintained replacement for what this node does, with more output formats, more control, and none of the deprecation baggage.

Why it's deprecated, and what to use instead

This is worth explaining rather than just repeating the warning: ADE_AnimateDiffCombine was written specifically as an output node for AnimateDiff workflows, scoped to a fixed set of formats and no real flexibility beyond GIF/webm-style output with basic ping-pong and loop-count options. Video Combine (from VideoHelperSuite) is a general-purpose video assembly node maintained as its own thing, with a much broader set of output formats and codec options, and it's what basically every current AnimateDiff tutorial and shared workflow actually uses. There's no functional reason to prefer the old node over the new one for anything you're building today.

The inputs and outputs that matter (if you inherit an old workflow using this)

Required: images (your generated frame batch), frame_rate (default 8, range 1–24), loop_count (default 0, range 0–100 - how many times to repeat before stopping, 0 meaning infinite for looping formats), filename_prefix (default "AnimateDiff"), format (a dropdown of supported output containers), pingpong (boolean - play forward then reverse for a cheap loop effect), and save_image (boolean, default true). Output is GIF, and this is also an output node in its own right (it writes the file to disk directly, same category as a Save Image node) - nothing needs to consume its output further downstream.

How to install it

If you're keeping an old workflow around that still uses this node, it's part of the same pack - no separate install:

cd ComfyUI/custom_nodes && git clone https://github.com/Kosinkadink/ComfyUI-AnimateDiff-Evolved

But if you're setting up video output fresh, install ComfyUI-VideoHelperSuite instead (ComfyUI Manager, search for it by name) and use its Video Combine node.

Common issues & troubleshooting

Should I migrate an old workflow off this node? Yes, whenever it's convenient - swap it for Video Combine, reconnect the images input, and you'll get more format options with no downside. There's no compatibility reason to keep the old node once you're touching the workflow anyway.

Frame rate capped at 24. That's a hard limit in this node's schema (max: 24) - another reason to prefer Video Combine if you want higher frame rates for smoother output or frame-interpolated clips.

GIF output looks low quality compared to what you see in tutorials. Most current tutorials aren't using this node at all anymore - they're on Video Combine with mp4/webm output, which looks noticeably better than GIF's limited color palette. That's less a setting to fix and more a reason to switch nodes entirely.