id: realvisxl-v5_0
name: RealVisXL V5.0
type: image-diffusion-model
task:
  - image-generation
language:
  - en
  - multilingual
source: huggingface
upstream_id: SG161222/RealVisXL_V5.0
upstream_url: https://huggingface.co/SG161222/RealVisXL_V5.0
license: openrail++
status: candidate
created_at: 2026-07-04
storage_path: /models/image/realvisxl-v5.0
format: diffusers
quantization: fp16
context_length: null
vram_required_gb: 12
parameters: null
base_model: stabilityai/stable-diffusion-xl-base-1.0
adapter_for: null
datasets: []
eval_suites: []
deployment:
  target: docker-gpu.cin.su
  runtime: transformers
  served_model_name: realvisxl-v5
notes: "Photorealistic SDXL finetune for local RTX 4090 image generation. The Hugging Face model card exposes Diffusers usage and targets photorealistic output."