stability ai

Stable Diffusion 1.5

Legacy

by Stability AI

The original widely-adopted image generation model. Extremely lightweight — runs on 4GB VRAM. Massive legacy ecosystem of checkpoints, LoRAs, and tools. Still preferred for speed and low VRAM scenarios.

VRAM requirements, GPU fit, and setup notes for Stable Diffusion 1.5, including 8GB/12GB fit guidance where relevant. Recommended runtimes: ComfyUI, Automatic1111, and Diffusers support. Best download size: ~2.0 GB at FP16.

  • Runs on 4GB+ VRAM — most accessible model
  • Massive ecosystem of fine-tuned checkpoints
  • Fastest generation of any quality model
  • Legacy but still widely used for speed
ComfyUI, Automatic1111, DiffusersFP16 safetensors

Your hardware

Detecting...

Parameters0.86B
Max Resolution512×512
Default Steps20
ArchitectureUNET
Licenseopenrail-m

Image Quality Benchmarks

Measured quality metrics for Stable Diffusion 1.5 outputs.

Human Preference Score50%

How often humans prefer this model's output (0-100%)

Aesthetic Score5.8

Visual quality and composition rating (5-9 scale)

CLIP Score0.26

Text-image alignment accuracy (higher is better)

VRAM Requirements by Resolution and Precision

Compare which GPUs can run Stable Diffusion 1.5 at different precisions. FP8 uses less memory than FP16 when available, and the grade shows how comfortably each GPU handles the workload.

FP16 (full precision)

ResolutionVRAM RequiredRTX 4090 24GBRTX 3060 12GBRTX 4060 8GBMacBook Pro M4 Pro 24GB
512×5124.0 GBSSSS
768×7686.5 GBSSAS
1024×10245.9 GBSSSS

Optimization Tips

Turbo / LCM distillation

Use distilled scheduler at 4-8 steps for faster iteration

ControlNets available

Add guided generation with 8 adapters (+0.7 GB VRAM each)

Rich LoRA ecosystem

Customize style, characters, and quality with community LoRAs

Run with Python

Run with Python (diffusers)
from diffusers import StableDiffusionPipeline
import torch

pipe = StableDiffusionPipeline.from_pretrained(
    "stable-diffusion-v1-5/stable-diffusion-v1-5",
    torch_dtype=torch.float16
)
pipe.to("cuda")

image = pipe(
    prompt="your prompt here",
    num_inference_steps=20,
    guidance_scale=7.5,
    height=512,
    width=512,
).images[0]
image.save("output.png")

Get started

Setup instructions for running Stable Diffusion 1.5 locally

1. Download the model

Get the checkpoint from HuggingFace

2. Place in:

ComfyUI/models/checkpoints/

3. Launch ComfyUI

python main.py

ComfyUI Workflow

Basic txt2img workflow for Stable Diffusion 1.5

7 nodes

Drag & drop into ComfyUI or use File → Import

Memory Breakdown

VRAM allocation at 1024×1024 on RTX 4090 24GB (24 GB)

Required: 5.9 GBAvailable: 24.0 GB
Weights1.7 GB
VAE0.2 GB
Text Encoder0.2 GB
Activations2.0 GB
Overhead0.5 GB

Estimated Generation Time

Time per image at 1024×1024, 28 steps, FP16.

RTX 4090 24GB~1s
RTX 3060 12GB~3.8s
RTX 4060 8GB~5.7s
MacBook Pro M4 Pro 24GB~8.1s

Sample Outputs

Available Formats, Downloads & Setup

Download Stable Diffusion 1.5 in the precision that matches your GPU. Lower precision usually means less VRAM pressure, while higher precision keeps more quality.

FormatPräzisionGrößeAnbieter
safetensorsEmpfohlenFP162.0 GBofficialHerunterladen
safetensorsFP324.0 GBofficialHerunterladen
ckptEmpfohlenFP162.1 GBofficialHerunterladen

ControlNet Support

8 ControlNets available for Stable Diffusion 1.5. ControlNets add guided image generation (edges, depth, pose) at the cost of extra VRAM.

Canny Edge

+0.7 GB VRAM

The original and most popular ControlNet. Edge-based structural guidance.

comfyuiautomatic1111diffusers
View on HF

Depth Map

+0.7 GB VRAM

Depth-based 3D spatial control.

comfyuiautomatic1111diffusers
View on HF

OpenPose

+0.7 GB VRAM

Human pose estimation for character control.

comfyuiautomatic1111diffusers
View on HF

Scribble

+0.7 GB VRAM

Rough sketch to image. Great for quick conceptual exploration.

comfyuiautomatic1111diffusers
View on HF

Lineart

+0.7 GB VRAM

Clean lineart extraction for illustration and coloring workflows.

comfyuiautomatic1111diffusers
View on HF

Normal Map

+0.7 GB VRAM

Surface normal estimation for material and lighting control.

comfyuiautomatic1111diffusers
View on HF

Tile/Upscale

+0.7 GB VRAM

Tile-based generation for upscaling and detail enhancement.

comfyuiautomatic1111diffusers
View on HF

Inpaint

+0.7 GB VRAM

Mask-based inpainting for editing specific regions of images.

comfyuiautomatic1111diffusers
View on HF

LoRA Ecosystem

Massive Ecosystem

The largest LoRA ecosystem in AI image generation. Thousands of LoRAs on CivitAI covering every imaginable style, character, concept, and quality modifier. SD 1.5 remains the most customizable image model.

Approximately 50,000 LoRAs available on CivitAI. Each LoRA adds ~0.1 GB VRAM.

Popular LoRAs for Stable Diffusion 1.5

NameCategoryDownloads
Detail Tweakerquality1.0MView
LowRAquality800KView
epiCRealismstyle600KView
Flat Color Animestyle500KView
Browse all LoRAs on CivitAI

Related Workflows

You might also like

Frequently asked questions

FAQ — Stable Diffusion 1.5 VRAM, Runtimes & Fit

How much VRAM does Stable Diffusion 1.5 need?

Stable Diffusion 1.5 (0.86B parameters) requires approximately 5.9 GB of VRAM at FP16 precision for standard 1024×1024 image generation. If you want a lighter setup, lower precisions like FP8 can reduce memory use when available.

Can I run Stable Diffusion 1.5 on an 8GB GPU?

Stable Diffusion 1.5 usually needs more than 8GB for comfortable local use. Check the VRAM table above for the exact resolution and precision trade-off.

Does Stable Diffusion 1.5 work in ComfyUI and Automatic1111?

Stable Diffusion 1.5 is marked for ComfyUI, Automatic1111, and Diffusers support in our catalog, so those are the runtimes we recommend first for local setup. If your workflow uses another front end, check the model's available formats and workflow notes above before downloading.

Can I run Stable Diffusion 1.5 on RTX 4090?

Yes, the RTX 4090 (24 GB VRAM) can run Stable Diffusion 1.5 comfortably at FP16. Expected generation time is around ~1s per image at 1024×1024.

Does Stable Diffusion 1.5 support ControlNet?

Yes, Stable Diffusion 1.5 has 8 ControlNet adapters available: Canny Edge, Depth Map, OpenPose, Scribble, Lineart, Normal Map, Tile/Upscale, Inpaint. Each ControlNet adds roughly 0.7 GB of extra VRAM.

Does Stable Diffusion 1.5 have LoRA support?

The largest LoRA ecosystem in AI image generation. Thousands of LoRAs on CivitAI covering every imaginable style, character, concept, and quality modifier. SD 1.5 remains the most customizable image model. The LoRA ecosystem for Stable Diffusion 1.5 is rated as "massive". There are approximately 50,000 LoRAs available on Civitai. Each LoRA adds roughly 0.1 GB of extra VRAM.

How fast is Stable Diffusion 1.5?

On a reference GPU (RTX 4090 24GB), Stable Diffusion 1.5 generates a 1024×1024 image in approximately ~1s at FP16 with 28 inference steps. Faster GPUs with higher memory bandwidth will produce images more quickly.

About Stable Diffusion 1.5

Use cases
artanimefast-generation
Recommended runtimes
comfyuiautomatic1111diffusers

See also