How to Generate Product Photos With AI for Free

Affiliate links ↓

Updated · May 12, 2026

Hiring a product photographer costs $150 to $800 per session in 2026, and that’s before retouching fees. That math breaks most early-stage product businesses before they’ve made a single sale. The better news: you can produce convincing, professional-looking product photos using three free tools, a smartphone, and about 30 minutes. This walkthrough covers the exact steps, the specific settings to use, and what to do when the AI renders your product as a melted prop from a sci-fi set.

What you need before starting:

• A product photographed on a clean surface — your kitchen table works fine
• Free accounts at Remove.bg, Leonardo AI, and Canva
• Time: about 30 minutes on your first run, 10 minutes once you’ve done it once

All three tools have genuinely functional free tiers for this workflow. The real limits — and honest workarounds — are covered at each step below.

1. Photograph your product on a plain background

Don’t overthink this. Place your product on a white sheet of printer paper or a white poster board. Natural window light is enough — no lightbox required. Shoot from eye level and at a 45-degree angle, take three or four shots, and keep the one with the sharpest focus on the product’s front face.

Two things actually matter here: use a plain white or light grey background (it makes the removal in step 2 faster and cleaner), and note which side the light is hitting your product. You’ll need to match that lighting direction in your AI-generated scene, or the composite will look immediately fake.

Avoid glossy tabletop surfaces that reflect background color back onto the product. A single sheet of matte white paper solves this.

2. Remove the background

Upload your product photo to Remove.bg. Drag the file into the upload box on the homepage — the background disappears in roughly three seconds. Click Download to get a transparent PNG.

The honest limitation: Remove.bg’s free tier exports at a maximum of 625×400 pixels. For social media or draft validation, that’s completely fine. For an Amazon or Etsy listing, you need higher resolution. In that case, use PhotoRoom‘s free web app instead — it removes backgrounds at higher resolutions and also lets you preview AI-generated scenes in the same interface. Free exports from PhotoRoom carry a watermark, though, which is why we use Canva for the final composite below.

If the removal cuts into your product edges — common with fine details or products close in color to the background — use PhotoRoom’s free manual brush tool to touch up the cutout before moving on.

3. Generate a backdrop with Leonardo AI

Leonardo AI’s free tier gives you 150 tokens per day, which works out to roughly 10 to 15 image generations at standard quality. That’s more than enough to land a usable background in one session.

1. Log in and navigate to Image Generation in the left sidebar.
2. Select Leonardo Phoenix as your model — it’s the most capable option available on the free tier as of mid-2026.
3. Write a prompt describing the scene environment, not your product. You’re generating a background, so leave the product entirely out of the description.
4. Set dimensions to match your target platform: 1:1 for Instagram, 4:5 for Amazon main images, 16:9 for banner use.
5. Click Generate and review the four outputs. Pick the one with the most consistent lighting and a natural focal point where your product will sit.

A concrete example: for a skincare serum, a prompt like “luxury bathroom countertop, white marble surface, soft morning window light, small eucalyptus sprig, neutral tones, photorealistic, editorial product photography“ produces usable results in one or two generations. The phrase “editorial product photography” steers the output away from the generic stock photo look.

If the scene feels too sterile, add small specific details: “a folded linen cloth in the background,” “condensation on a nearby glass,” “a few loose dried botanicals.” Specificity consistently beats vagueness in image prompts — this applies across every AI image tool, not just Leonardo.

4. Composite your product into the scene with Canva

Open Canva and create a new design at your target dimensions. Upload both files — the AI-generated background and your transparent product PNG from step 2.

1. Click the background image and select Set as background to lock it as the base layer.
2. Upload your transparent product PNG, click Add to design, and drag it into position in the scene.
3. Resize the product so it looks proportionally correct in the environment. The most common mistake is leaving it too large — it reads as pasted on rather than present in the space.
4. Add a ground shadow: insert a circle or ellipse shape element from the Elements panel, flatten it horizontally beneath the product, and pull the transparency slider to around 25%. This single step is what separates a composited-looking image from a finished one. Skipping the shadow is the most frequent tell.
5. Select your product layer and open Edit image to nudge brightness and contrast. A -5 to -10 brightness adjustment is usually enough to visually ground the product in the scene’s lighting.

Export via Share → Download → PNG. Canva free accounts include full-resolution PNG export for designs you’ve created yourself, with no watermark.

What do you do when the composite looks wrong?

Three issues cause most failed composites: mismatched lighting direction, a missing drop shadow, and scale that’s slightly off. Match the lit side of your product to the scene’s light source, add a flattened ellipse at 25% opacity beneath the product, and resize until the proportions feel natural. Most bad composites are fixable with those three adjustments alone.

The background removal cut into your product. Re-shoot on a higher-contrast background — dark product on white, light product on black. Alternatively, use PhotoRoom’s manual brush tool to paint back the edges. The brush is available in the free web app under the Edit panel after removal.

Leonardo AI ran out of tokens mid-session. Tokens reset daily at midnight UTC. In the meantime, Canva’s built-in Text to Image generator — found in the Elements panel — produces backgrounds for free at lower quality. For simple surfaces like marble countertops or wooden tables, it’s a workable fallback. ChatGPT‘s free image generation is another option with the same prompt approach; slightly less photorealistic for product scenes but reliable for basic backdrops.

The lighting mismatch looks too extreme to fix in Canva. Go back to Leonardo and regenerate a few backgrounds with an explicit lighting direction in the prompt: “light source from the left, strong directional shadows.” Match to your product shot specifically. It’s faster to generate a new background than to correct a lighting mismatch in post.

How do you scale this beyond a handful of products?

If you’re processing more than 20 SKUs, the manual Canva composite step becomes your real bottleneck. PhotoRoom’s paid plan (around $10/month) handles batch processing across your full catalog, applying background removal and AI scene placement in bulk. The time math shifts quickly once you’re above 20 products per month.

For a solo seller or small team staying on the free workflow, there’s a useful shortcut: once you’ve found a Leonardo scene prompt that works well for your brand, save it and reuse the exact same prompt for every product. Consistent AI backgrounds across a product line read as intentional visual branding, not laziness. You’ll also get more predictable results because you’ve effectively tuned the prompt already.

One exception worth knowing: products with reflective surfaces — glass bottles, metallic packaging, glossy electronics — are hard to composite convincingly because real reflections in the product surface won’t match the AI background. For those, a simple DIY lightbox (cardboard box, white tissue paper, two desk lamps) shot against a real environment will outperform the AI composite workflow on quality, even with a smartphone.

Frequently asked questions

Are AI-generated product photos allowed on Amazon and Etsy?

Yes, as of 2026 both platforms permit AI-enhanced product images as long as they accurately represent the actual product. You can’t substitute a fabricated product image, but AI-generated backgrounds, lighting, and scene composition are allowed. Check each platform’s current listing policies directly before uploading, since these rules have been updated in the past 18 months.

How do I match the AI background lighting to my product photo?

Include the lighting direction explicitly in your Leonardo prompt: “light source from the left” or “soft overhead diffused lighting.” Then in Canva, adjust the product layer’s brightness slightly and make sure the ellipse shadow falls in the same direction as the scene light. If the gap is still obvious, regenerating two or three background variations is usually faster than trying to color-correct your way out of a bad match.

Can I do this entirely inside one tool instead of three?

Yes — PhotoRoom handles background removal and AI scene generation in a single interface, which cuts the workflow down to about five minutes per product. The trade-off is that free exports carry a watermark. For testing scenes and validating that a visual direction works before committing to a paid plan, it’s genuinely useful as a starting point.

Related reads

• Best AI Tools for Podcast Editing in 2026
• Canva AI vs Adobe Firefly: Which Is Better for Designers?
• Best AI Tools Under $20 a Month: What’s Worth It

This article contains affiliate links. If you subscribe through one, we may earn a commission at no extra cost to you. It never changes what we recommend — we only link to tools we actually use. Full disclosure.