Upload Image
Give Animon a local image file and get a server-side IMAGE_ID
- animon_key
- IMAGE_ID
The node that bridges your files and the cloud
The Animon nodes don't see your hard drive. When Start-End to Video asks for ANIMON_IMAGE_IDs, it's asking for files that are already sitting on Animon's server - and Animon Upload Image is how files get there. It works the way you'd expect LoadImage to work: a dropdown lists the images in ComfyUI's input folder, you pick one, and the node uploads it. The difference is what comes out.
LoadImage hands you an IMAGE tensor for local nodes. This node hands you an IMAGE_ID - a handle Animon can reference in a future generation call. Same source file, completely different purpose.
How it works
The node reads the list of image files from folder_paths.get_input_directory() (ComfyUI's input folder, the same place LoadImage looks), filters to image content types, and shows them in the image combo. On execution it opens the file and calls Animon's files.create(file=..., purpose="video_generation") - the OpenAI-style Files API, with the upload tagged for video work. The returned file ID is your IMAGE_ID output, ready to feed Start-End to Video's image_start and image_end inputs.
One input besides the key: the image combo itself. If the file you want isn't in the list, drop it into ComfyUI/input/ and refresh the node - same muscle memory as LoadImage.
Installing and when you'd actually use it
Install is pack-standard: ComfyUI Manager → search ComfyUI-Animon → restart, then thread a Key node in. Requirements are just openai and decord.
You'll specifically need this node when you have a file on disk that isn't part of your ComfyUI graph - say a reference frame you drew or sourced elsewhere. If the image came out of your own diffusion pipeline as a tensor instead, Upload Image (from Tensor) is the better fit and skips the file round-trip. But for the classic "pick a starting frame from my input folder" flow, this is the node. Upload once, reuse the IMAGE_ID as many times as you like - the README even suggests reusing an ID rather than re-uploading the same image when your start and end frames match.
Inputs (2)
| Name | Type | Default | Description |
|---|---|---|---|
| animon_key | ANIMON_KEY | — | |
| image | COMBO | 1 options: example.png |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| IMAGE_ID | ANIMON_IMAGE_ID | — |