How to Upload Images & Use the AI Design Studio

Overview

Great ad creative is the single biggest driver of campaign performance. Agency AI gives you three ways to source images for your Image ads: upload your own, use AI-powered tools to edit or generate images from scratch, and pull from your Asset Library. This guide walks you through each option so you can create scroll-stopping ads without needing a graphic designer.

Accessing the Image Tools

The image section appears at the bottom of the ad creation screen when you select Image as your Ad Type. You'll see three tabs:

Upload — Upload images directly from your device
AI Tools — Edit existing images or generate new ones with AI
Asset Library — Access saved images (coming soon)

Tab 1: Upload

The Upload tab is the quickest way to add images to your ad.

How to upload:

Click the Upload tab (selected by default).
Click Add images in the upload area.
Select up to 2 images from your device.
Once uploaded, use the checkboxes to select which images to include in your ad.

File requirements:

Accepted formats: .jpg or .png
Maximum file size: 30 MB

Image size recommendations:

For best results, upload one Square (1:1) image optimized for Feed placements (Facebook Feed, Instagram Feed) and one Vertical (9:16) image optimized for Stories and Reels placements.
If you only have one image, that's perfectly fine — Agency AI will automatically optimize it to fit all placements.

Tips for great ad images:

Use high-resolution images that look sharp on mobile devices. Most people will see your ad on their phone.
Lifestyle imagery showing your product in use tends to outperform plain product shots on white backgrounds.
Keep text on images to a minimum. Meta's algorithm can penalize ads with too much text in the image, which reduces delivery.
Make sure your product is the clear focal point of the image. Avoid cluttered or busy backgrounds that distract from what you're selling.

Tab 2: AI Tools

The AI Tools tab is where Agency AI's creative capabilities really shine. It offers two modes: Edit an existing image or Generate a brand new one from a text prompt.

AI Image Editor (Edit Mode)

The AI Image Editor lets you modify an existing image using natural language instructions. This is useful for adjusting images you already have without needing photo editing software.

How to use the AI Image Editor:

Click the AI Tools tab.
Select the ✏️ Edit option (selected by default).
Under AI Image Editor, click Add image and upload the image you want to edit (.jpg or .png).
In the Edit instructions field, describe what you want to change. Write your instructions in plain language — for example, "Remove the background and replace it with a clean white studio backdrop" or "Make the colors more vibrant and add warm lighting."
Select your Image size from the dropdown. Choose Feed for a square (1:1) image optimized for Feed placements, or select the appropriate size for your target placement.
Click Process Edit to apply your changes.
If the result isn't what you wanted, click Start over to reset and try again with different instructions.

Tips for effective edit instructions:

Be specific about what you want changed. "Make it better" is vague — "Brighten the lighting and increase the contrast" gives the AI a clear direction.
Focus on one or two changes at a time for the most accurate results.
Reference specific areas of the image when possible, such as "Remove the object in the background" or "Change the shirt color to navy blue."
If you're not happy with the first result, try rephrasing your instructions. Different wording can produce meaningfully different outcomes.

AI Image Generator (Generate Mode)

The AI Image Generator creates brand new images from a text description. This is ideal when you don't have product photography, want to create lifestyle scenes, or need creative concepts that don't exist yet.

How to use the AI Image Generator:

Click the AI Tools tab.
Select the ✨ Generate option.
Under AI Image Generator, enter your image description in the Prompt field. Describe the image you want to create in as much detail as possible — for example, "A woman wearing a colorful tie-dye hoodie at a music festival, golden hour lighting, crowd in the background, vibrant and energetic mood."
Select your Image size from the dropdown. Choose Feed for a square (1:1) image or select the appropriate size for your target placement.
Click Generate image to create your image.

Tips for writing effective prompts:

Include details about the subject, setting, lighting, mood, and style you want.
Mention the type of product if it should appear in the image, along with how it's being used or worn.
Specify the visual tone — words like "bright," "moody," "minimal," "energetic," or "luxurious" help steer the output.
If you want a specific composition, mention it — for example, "close-up shot," "overhead flat lay," or "wide angle with space for text overlay on the left."
Generate multiple versions with slightly different prompts and test them against each other in your ads to see which creative resonates best.

Tab 3: Asset Library

The Asset Library is a centralized place to store and reuse images across your ad campaigns. This feature is coming soon and will allow you to save your uploaded and AI-generated images for quick access when creating future ads.

Once available, the Asset Library will eliminate the need to re-upload or regenerate images you've already used, making it faster to spin up new ad variations.

Choosing the Right Image Strategy

If you have strong product photography: Use the Upload tab to add your best images directly. Consider using the AI Image Editor to create placement-specific versions (e.g., editing a landscape photo into a vertical crop for Stories).

If your product photos need improvement: Upload them through the AI Tools > Edit tab and use instructions to enhance lighting, remove backgrounds, or adjust the composition for ad placements.

If you don't have product photography: Use the AI Tools > Generate tab to create images from scratch. This is especially useful for new stores that haven't invested in a photoshoot yet, or for creating lifestyle and mood-based imagery that complements your existing product shots.

For testing multiple creatives: Use a combination of all three approaches. Upload your best existing image, generate an AI alternative, and test both in the same campaign using the Create New Ad option to see which creative performs better.

Image Size and Placement Guide

Meta delivers ads across multiple placements, and different placements have different ideal aspect ratios:

Feed placements (Facebook Feed, Instagram Feed, Marketplace) — Square (1:1) images work best. This is the most common placement and where most of your impressions will likely come from.

Stories and Reels placements (Facebook Stories, Instagram Stories, Instagram Reels) — Vertical (9:16) images fill the full screen and feel native to the format. Using a properly sized vertical image for these placements significantly improves engagement.

Right Column (Facebook desktop only) — Uses smaller image sizes and is a minor placement for most advertisers.

By uploading or generating both a 1:1 and a 9:16 version of your creative, you ensure your ad looks its best everywhere Meta places it. If you only provide one image, Agency AI will automatically adapt it for different placements — but purpose-built images for each format will always outperform auto-cropped versions.

Need Help?

If you have questions about image requirements, AI tools, or which creative approach is right for your store, contact our support team. We're happy to help you get the best possible creative for your campaigns.