Quick concept exploration
type prompt, click generate, get result
// official site: github.com ↗
AI / LLM · PRO TIER
Fooocuspro
Fooocus is a Midjourney-style easy-mode image generator built on top of Stable Diffusion — designed by lllyasviel (the creator of ControlNet) to deliver "first generation in 30 seconds with zero configuration". Hand-tuned defaults + smart prompt enhancement make it the friendliest entry point to SD.
// What it is
A closer look.
Fooocus is a Midjourney-style easy-mode image generator built on top of Stable Diffusion — designed by lllyasviel (the creator of ControlNet) to deliver "first generation in 30 seconds with zero configuration". Hand-tuned defaults + smart prompt enhancement make it the friendliest entry point to SD.
For users who don't want to learn ComfyUI nodes or A1111 settings panels, Fooocus is the answer.
// Use cases
What it's for.
Concrete scenarios where teams pick Fooocus over the SaaS alternative.
Photo-realistic + artistic styles
out-of-the-box
Image prompting
drop reference image, get variations
Inpainting + outpainting
minimal-friction edit workflows
Advanced controls
available but hidden by default ("Advanced" toggle)
LoRA chains + ADetailer
auto-applied behind the scenes
// Who it's for
Built for these teams.
If your team profile matches one of these, Fooocus is a strong fit out of the box.
Non-technical users
designers, marketers, content creators
Beginners
to AI art exploring possibilities
Quick brainstorming
concept exploration before serious A1111/ComfyUI work
Photo studios
offering quick generation tasks to clients
Hosting providers
selling consumer-friendly AI image tier
// Differentiators
Why teams pick Fooocus.
When evaluating self-hosted options for this category, here are the dimensions on which Fooocus consistently lands above the alternatives.
- ✓GPL-3.0 — fully open
- ✓Minimal UI — radically simpler than A1111 / ComfyUI
- ✓Smart defaults — created by ControlNet author, knows what good results look like
- ✓Style presets — Cinematic, Anime, Photorealistic, etc. one-click
- ✓Image prompt — paste reference for variation (instead of prompt engineering)
- ✓30 sec first result — by design, friendly cold-start
// Integrations
Connects to.
The stack you'll plug Fooocus into — services, protocols, and adjacent apps in the BluixApps catalog.
◇
SDXL base + refiner
proper SDXL workflow auto-configured
◈
LoRA support
UI hides complexity, just toggles
◆
ControlNet
under "Advanced" but works
▣
Image prompting
drag-drop ref into prompt area
▦
API mode
via mrhan1993/fooocus-api Docker image (BluixApps default)
▩
Inpaint / outpaint
modal canvas similar to InvokeAI
▼
Reverse proxy
HTTPS + auth for production
// Adoption & deployment
Notable users & community
- 41k+ GitHub stars
- Used as default SD tier on many cloud platforms
- Tutorial-friendly (lots of YouTube content)
- Featured in Stability AI release demos
- Strong on consumer / hobby creators side
What we ship
- Docker (mrhan1993/fooocus-api:latest) with NVIDIA runtime + API mode
- Persistent volumes for models + outputs
- Port 7865 mapped
- Install report at
/root/bluixapps/fooocus.txt - Pre-configured
--listen --port 7865 - Backup hook covers models + outputs
- Pairing note: ComfyUI for advanced when Fooocus simplicity isn't enough
- GPU pre-flight check via
bluixapps_ensure_nvidia_runtime
// Tips & operations
Run it properly.
Operational guidance from running this in production — what to lock down, what surprises people.
// PERFORMANCE
VRAM
4-6 GB enough for SD 1.5; 8GB for SDXL; 12-16 GB optimal
// SECURITY
First gen
pre-pulls SDXL base + refiner (~12 GB) — let it complete
// OPERATIONS
Style presets
try "Fooocus V2" + "Cinematic Default" for best photoreal
// RELIABILITY
Image prompt
way more powerful than text alone — always use when possible
// DEPLOYMENT
API mode
BluixApps uses mrhan1993 fork that exposes REST endpoints
// SCALING
Persistent paths
/content/Fooocus/models (LoRAs, checkpoints), /content/Fooocus/outputs
// MAINTENANCE
Production
reverse proxy + basic auth + HTTPS
8192
// min ram (MB)
30
// min disk (GB)
7865
// access port
http
// protocol
pro
// bluixapps tier
// Alternatives in AI / LLM
Compare with
Project resources
Official sitegithub.com ↗