Stable Diffusion 1.5 Resolution Calculator

Stable Diffusion 1.5 was trained at 512×512, and its VAE works in multiples of 8. Staying near 512×512 worth of pixels (262,144) avoids duplicated subjects, so for other aspect ratios the best size keeps the total close to that budget. Pick a ratio to get the best size and the exact crop to finish on.

Jump to the FAQ ↓

Stable Diffusion 1.5 · grid model · last verified 2026-10-06 · Source

Model

Stable Diffusion 1.5 · change model

Aspect ratio

Type W:H, W/H or a decimal. Flip swaps portrait and landscape.

Need a GPU to run this?
Rent on RunPod →Rent on Vast.ai →

Affiliate links: I earn a commission if you sign up, at no cost to you.

Full table

FAQ

SD 1.5 size questions

What is the best Stable Diffusion 1.5 size for 16:9?

680×384 is the largest size within 1% of 16:9 that stays inside the 512×512 pixel budget. 640×360 is the most exact 16:9 if you want zero error. Both are multiples of 8.

Why does Stable Diffusion 1.5 repeat subjects at large sizes?

The model was trained at 512×512. Sizes with far more pixels than that can lead to duplicated subjects, so this tool keeps the total near 262,144 pixels.

Does the size have to be a multiple of 8?

Yes. The VAE downsamples by a factor of 8, so widths and heights need to be multiples of 8.

How big can I print a Stable Diffusion 1.5 image?

At 300 PPI, 512×512 prints at about 1.7×1.7 in. A 4× upscale to 2048×2048 prints at about 6.8×6.8 in. Check other sizes.

Other models
SDXLFLUX.1Qwen-ImageMidjourneyFLUX.2FLUX 3 ImageGPT Image 2 / 2.5
Printing an AI image? Check how big it will print. Print size →