Convert AI ad creatives to JPG or PNG

Prompting is now the fastest way to design an ad. Claude Design, ChatGPT, Gemini and app builders like v0 or Lovable turn a brief into a finished creative in minutes, complete with gradients, product shots and bold typography. There is just one catch: the output is HTML, and Meta, Google Ads, LinkedIn and TikTok only accept image files. This converter renders your HTML creative in a real Chromium browser and exports it as a high-resolution JPG or PNG, ready to upload to any ad platform.

Drop your .html or .zip hereor click to browse. Up to 50 MB, no watermark

Your file is processed in an isolated browser and deleted immediately after conversion.

The gap between AI ad tools and ad platforms

Ad platforms want JPG or PNG at specific dimensions. An HTML file cannot be uploaded to Meta Ads Manager, Google Ads or LinkedIn Campaign Manager.
Manual screenshots never match the exact ad dimensions, are capped at your monitor's resolution and catch scrollbars and browser chrome.
Rebuilding the creative in Canva or Figma just to export an image throws away the speed that made prompting attractive in the first place.
Print-to-PDF-then-crop workflows go through the print pipeline, which turns gradient text black and strips shadows before you ever reach an image.
A/B testing means many variations, and capturing ten creatives by hand, one by one, does not scale.

How to convert

1.Design the ad in your AI tool at the platform's dimensions (for example a 1080x1080 frame for a Meta feed ad, 1080x1920 for a story).
2.Export it as HTML. In Claude Design: Share, then "More formats and apps", then HTML. From ChatGPT or Gemini, download the file or save the code as .html.
3.Upload the file (or the exported .zip, as is) above and pick JPG or PNG in the Format option.
4.The engine renders the creative in headless Chromium at 2x resolution. One creative downloads as a single image; a file with several variations downloads as a zip of numbered images.

Works with any tool that outputs HTML

Claude Design is the flagship use case: describe the ad in a prompt, refine it with Edit and Annotate, export as HTML and convert here. But the same flow works with anything that produces HTML: Claude Artifacts, ChatGPT (Canvas or a plain code block), Gemini, Copilot, v0, Lovable, Bolt, or a creative you hand-coded.

Dedicated ad generators that already export finished images do not need a converter. Renditions is for the growing number of teams who design ads by prompting a general AI, because that is where iteration is fastest, and who then need a file the ad platform will actually accept.

Hit each platform's specs

The capture is always 2x the CSS dimensions of your design, so the way to hit a platform spec is to design at that size. A 1080x1080 frame becomes a 2160x2160 image, which Meta and Google Ads accept happily since they only enforce aspect ratio and minimum size. Need several formats of the same ad? Ask your AI tool for a square, a landscape and a story variant in one file, and Renditions will capture each one.

Variations at scale

A/B testing lives on volume: same offer, five headlines, three color ways. Generate all the variations in a single HTML file, one full-frame section per creative, and convert it once. You get a zip with every variation as a numbered image (slide-01.jpg, slide-02.jpg…), captured in order at full resolution. One prompt, one upload, a complete ad set.

FAQ

Which AI tools does this work with?

Any tool whose output is HTML: Claude Design, Claude Artifacts, ChatGPT, Gemini, Copilot, v0, Lovable, Bolt or hand-written code. If the creative opens in a browser, it converts. Tools that already export a finished JPG or PNG do not need this step.

How do I export an ad from Claude Design as a JPG?

In Claude Design, open Share, choose "More formats and apps" and export as HTML. Upload the downloaded file here without unzipping, pick JPG in the Format option and convert. There is a dedicated Claude Design page with the full flow.

What dimensions will the image have?

Twice the CSS dimensions of your design, which keeps text and gradients sharp. Design at the platform's spec (1080x1080 for Meta feed, 1200x628 for Google display, 1080x1920 for stories) and upload the 2x capture as is; platforms accept larger files at the right aspect ratio.

Should I pick JPG or PNG for ads?

JPG for almost all ad platforms: encoded at quality 92 with full chroma resolution, it looks identical to the lossless capture and is several times smaller. Pick PNG when a platform requires it or for flat-color designs with razor-sharp small text.

My file contains several ad variations. What do I get?

A .zip archive with one numbered image per variation, captured in order. A single-frame file downloads as a plain .jpg or .png.

Is there a watermark or a cost?

No watermark, ever. Conversions are free every day; Premium at $6/month removes the daily limit for teams shipping ad sets constantly.

Free every day, no watermark. Converting a lot? Premium is unlimited for $6/month.

See pricing