COMFYDOCK

// local-first image generation

ComfyUI without the noodles.

Pick a workflow, click run. ComfyDock auto-installs missing models into the right folders and hides the node graph behind a simple form. Runs 100% on your machine.

Get started →↓ Download agent
connection status
…—
Local agentoffline
ComfyUI engineoffline
Paired devicenone
manage devices →
Capabilitiesn/a
Agent not running. See setup.

Workflow gallery

+ new custom workflow19 / 19
SDXL

SDXL — Text to Image

Classic SDXL base text-to-image. Good defaults at 1024×1024.

txt2imgOpen →
FLUX

Flux.1 schnell — Text to Image

Fast, sharp results from Flux.1 schnell in just 4 steps.

txt2imgOpen →
IMG2IMG

SDXL — Image to Image

Restyle or refine an existing image with SDXL.

img2imgOpen →
INPAINT

SDXL — Inpaint

Paint a mask, describe the fill. Best with an SDXL inpaint checkpoint.

inpaintOpen →
UPSCALE

4× Upscale — UltraSharp

Tile-friendly ESRGAN 4× upscale of any input image.

upscaleOpen →
SD1.5

SD 1.5 — Text to Image

Lightweight classic. Great on low-VRAM PCs (~2 GB model).

txt2imgOpen →
TURBO

SDXL Turbo — 1-step

Near-instant previews. 1–4 steps, CFG 1, 512px.

txt2imgOpen →
FLUX-DEV

Flux.1 dev — Text to Image

Higher quality Flux variant, 20 steps. Slower but sharper than schnell.

txt2imgOpen →
LIGHTNING

SDXL Lightning — 4-step

ByteDance Lightning LoRA on SDXL base. 4 steps, CFG 1.

txt2imgOpen →
FLUX-I2I

Flux — Image to Image

Restyle an image with Flux schnell. Lower denoise = closer to original.

img2imgOpen →
STYLE

SDXL — Style Transfer (IP-Adapter)

Lift the style of a reference image onto a new prompt. Needs IP-Adapter custom node.

img2imgOpen →
FLUX-FILL

Flux Fill — Inpaint

Cleaner inpaint using the Flux Fill dev model. Paint a mask, describe the fill.

inpaintOpen →
ANIME-2X

2× Upscale — RealESRGAN Anime

Crisp 2× upscale tuned for anime/illustration art.

upscaleOpen →
FACE

Face Restore — GFPGAN

Restore blurry or low-res faces. Needs ReActor/Facerestore custom node.

upscaleOpen →
TILE-2X

Tile Upscale — SDXL 2×

Detail-adding 2× upscale: pixel upscale → light SDXL refine.

upscaleOpen →
ANIMATE

AnimateDiff — SD 1.5

Short prompt-driven animation (16 frames @ 8 fps) using SD 1.5 + AnimateDiff v3.

txt2imgOpen →
SVD

Stable Video Diffusion — Image to Video

Upload an image, get a 25-frame clip. SVD XT 1.1, native nodes.

img2imgOpen →
LTX-V

LTX-Video — Text to Video

Lightricks LTX-Video 0.9.5. Fast txt→video, 768×512, ~5s. Native nodes.

txt2imgOpen →
WAN-2.1

Wan 2.1 — Image to Video (14B)

High-quality img→video using Wan 2.1 I2V 14B (fp8). Slow but stunning.

img2imgOpen →