AI Prompt for Image Workflows (ComfyUI, ControlNet)
A node-graph-level ComfyUI workflow specification to achieve turn pencil sketch into finished painting using Animagine XL 3.1 with Tile upscaling control.
Replace the bracketed placeholders with your own context before running the prompt:
[see positive prompt below]— fill in your specific see positive prompt below.[see negative prompt below]— fill in your specific see negative prompt below.[... continue for all nodes ...]— fill in your specific ... continue for all nodes ....[tuned for Animagine XL 3.1, 30-60 words]— fill in your specific tuned for animagine xl 3.1, 30-60 words.[standard negatives + checkpoint-specific]— fill in your specific standard negatives + checkpoint-specific.[e.g., DPM++ 2M Karras, Euler a, DDIM]— fill in your specific e.g., dpm++ 2m karras, euler a, ddim.[karras, normal, exponential]— fill in your specific karras, normal, exponential.[20-40 typical]— fill in your specific 20-40 typical.More prompts for Image Workflows (ComfyUI, ControlNet).
A consistent-character workflow generating fox in an enchanted forest of glowing mushrooms across poses using IP-Adapter FaceID Portrait and Scribble with Juggernaut XL v10.
A consistent-character workflow generating fox in an enchanted forest of glowing mushrooms across poses using IP-Adapter Composition and Depth (MiDaS) with RealVisXL V4.0.
A consistent-character workflow generating fox in an enchanted forest of glowing mushrooms across poses using IP-Adapter Composition and Shuffle with Pony Diffusion XL.
A consistent-character workflow generating astronaut standing on Mars cliff edge across poses using IP-Adapter Plus (style only) and Lineart anime with Animagine XL 3.1.
A consistent-character workflow generating astronaut standing on Mars cliff edge across poses using IP-Adapter FaceID Plus v2 and Canny edge with SDXL 1.0 base.
A node-graph-level ComfyUI workflow specification to achieve remove background cleanly with hair detail using Flux.1 [dev] with Tile upscaling control.