Convert aspect ratios to pixels, choose export sizes for social platforms, and copy Midjourney --ar flags for image and video prompts. This version includes current platform presets and safer model-size logic.
| Placement | Platform | Pixels | Ratio |
|---|
AI image models usually generate from a limited set of sizes, while Instagram, YouTube, TikTok, ads, print, and wallpapers each expect different final dimensions. This tool helps you start with the closest ratio before you spend credits.
If you generate a square image for a vertical Story or a wide thumbnail, important subjects can be cut off later. Pick the target platform first, then generate near that shape and resize only at the end.
Use the calculated aspect ratio for Midjourney, the recommended pixel size for image editors, and the nearest supported size for OpenAI, DALL-E, Stable Diffusion, Flux, or AI video workflows.
--ar, DALL-E 3 legacy for OpenAI Image API legacy sizes, or Stable Diffusion/Flux for diffusion-friendly multiples.Goal: a 16:9 YouTube thumbnail. Select YouTube thumbnail HD or YouTube thumbnail 4K, then choose DALL-E 3 legacy. The closest DALL-E 3 landscape generation size is 1792 x 1024px; export the final thumbnail as 1280 x 720px or 3840 x 2160px.
{
"model": "dall-e-3",
"prompt": "A clean 16:9 YouTube thumbnail with a centered subject and readable title space",
"size": "1792x1024"
}Goal: a vertical 9:16 mobile canvas. Select Instagram Story / Reel, YouTube Shorts cover, or TikTok in-feed vertical, then choose DALL-E 3 legacy. Generate at 1024 x 1792px, then export at 1080 x 1920px.
{
"model": "dall-e-3",
"prompt": "Vertical product poster, subject centered, empty space at top and bottom for app UI",
"size": "1024x1792"
}Select Instagram feed portrait, click Use this preset, then copy --ar 4:5. Generate with the subject centered and export the final file at 1080 x 1350px.
Select Pinterest standard Pin, then choose Stable Diffusion XL / SD3 or Flux / local diffusion. The calculator rounds dimensions to practical multiples so local pipelines are less likely to reject the size.
Platform specs change. These presets were reviewed against official or primary help pages where available, plus practical defaults for AI image workflows.
Start with the final placement whenever you know it. If you are generating in Midjourney, copy the whole-number --ar flag, generate close to the target, then crop or upscale to the exact platform preset. If you are using OpenAI Image API or DALL-E, choose the nearest supported model size first, then export the exact platform dimensions after generation.
Last reviewed: April 28, 2026. Verify exact requirements in the source platform before running paid campaigns.
Midjourney aspect ratio flags use whole numbers. Convert 1.39:1 to 139:100 before using --ar.
For GPT Image models, use the documented square, landscape, portrait, or auto size options. For DALL-E 3 legacy, use the closest of 1024x1024, 1792x1024, or 1024x1792, then resize or crop to the exact platform export.
Platform sizes are final delivery targets. AI generator sizes are model constraints. Generate near the right aspect ratio first, then crop, resize, or upscale.
Use 9:16. Practical exports are 1080x1920 for common vertical video and 1440x2560 when you want a higher-resolution working master.
No. It only calculates ratios and dimensions in your browser.