Uncensored image and video in ComfyUI, without a GPU

1. 1

   Install the node pack

   Open Manager in ComfyUI, search for SpicyAPI, click Install and restart ComfyUI. No Manager? Clone the pack into custom_nodes with the command below. A SpicyAPI group appears in the node menu, sorted into Image, Video, Audio, Text and Utils.

   cd ComfyUI/custom_nodes
   git clone https://github.com/spicyapi-ai/comfyui-spicyapi

   CopyRead the setup guide
2. 2

   Paste your API key

   Open Settings, then SpicyAPI, and paste a key from your console. A notice confirms it and shows your balance. The key is stored in ComfyUI's user folder, never in a workflow or in the metadata of a saved image, so sharing either is safe. Restart once so the node list comes from your account.

   
3. 3

   Find an uncensored model

   Double-click the canvas and type a model name. Nodes are named exactly like the models on our site, and the preview panel shows what each one does and what it costs. Spicy models are the uncensored tunes.

   
4. 4

   Chain image into video

   Connect a text-to-image node's images output to an image-to-video node's image input, then add a Save Video node. Every option on a node comes from the model itself, so a node never offers a setting the model would reject.

   
5. 5

   Run it and keep an eye on the price

   The badge on each node shows its rate for the options you picked. Every run is quoted first, and a Max cost per run in Settings stops anything above your limit before it is charged. When the node finishes it prints what it cost and the task ID. Seeds start on fixed, letting ComfyUI reuse unchanged node outputs while its execution cache is available. Running an uncached node starts a new billed task.

Prefer to start from a working graph? Open ComfyUI's template browser: under comfyui-spicyapi you get text to image, image to video, a chat model writing the prompt, text to speech, and a LoRA example that runs as is.