Tell us your hardware budget and what you want to make. We pick a consistent, working stack from our catalogue — workflow engine, model, runner, training tools. Empty slots tell you why nothing fits, instead of pretending.
Local video diffusion — Wan, Hunyuan, LTX, CogVideoX. The heaviest workload. Cards in this tier: RTX 3060 12GB, RTX 4070.
The nodal workflow engine for serious diffusion.
Role: The graph / UI that actually runs your pipelines. Where you build the model chain.
ALT // ComfyUI-Manager — install, update, and govern comfyui custom nodes.
Reference-image conditioning for ComfyUI.
Role: The image model itself — SDXL, Flux, SD3, etc.
ALT // FLUX.1 [dev] — 12b parameter open-weight diffusion model.
GPU-accelerated upscaling, frame-interp, denoise.
Role: Upscaling, interpolation, and post-processing.
ALT // FaceFusion — the most active open face-swap toolkit.
All the ControlNet preprocessors in one node pack.
Role: Plugins that expand the workflow engine — controlnet, ipadapter, custom samplers.
ALT // ComfyUI Impact Pack — detection, segmentation, and detailers — the canonical face/hand fix pack.
Every pick is a function of three things in our catalogue: minimum VRAM (must fit your budget), power-user score (60% of the weight), and trending score (20%). We add small bonuses for open-source licensing and beginner-friendly setup. No paid placements, no “sponsored” tier — if it's not in our catalogue, it can't appear here.
