Product Photography with AI: A Step-by-Step Workflow

To get product photography with AI, connect Polaris to Claude, attach your existing product photo in chat, describe the scene or shot you want, pick settings from the tappable option cards, and let the live panel render the finished image.

What you need

• At least one usable photo of the product — Polaris regenerates and reimagines it, it doesn't invent the product from nothing.
• A sense of the shots you want: white background, in-use, lifestyle, close-up detail, or flat lay.
• The Polaris MCP connector added to Claude, a one-time setup.

Step-by-step: generating product shots

1. Add the connector. In Claude, connect Polaris at polarisugc.com/api/mcp and authorize it once.
2. Attach your product photo. Upload the image in chat and describe the shot, e.g. "turn this into a clean white-background product shot" or "show this on a kitchen counter in morning light."
3. Choose settings from the option cards. Claude presents tappable cards for image model, aspect ratio, and style rather than requiring you to type technical parameters.
4. Watch the live render. A Polaris panel appears in the conversation, showing progress and then the finished shot, usually within 20 to 180 seconds.
5. Batch multiple shots. Ask for several angles or scenes at once so they render in parallel rather than one request at a time.
6. Download the set. Pull individual images or the full ZIP once the batch completes.
7. Iterate with Recreate. If a shot is close but the lighting or angle is off, use one-click Recreate rather than re-describing everything from scratch.

Pro tips

• Be specific about lighting and surface (e.g. "soft morning light on a wood table") — vague prompts produce generic-looking shots.
• Generate a mix of shot types in one batch (white background, lifestyle, close-up) so you have options for different placements: listing, ad, or social.
• Use the same source photo across a batch for visual consistency across your product page or ad set.
• Failed renders auto-refund, so it's low-risk to batch-test a few scene ideas before settling on one.

Honest caveats

Because Polaris works from your real product photo, quality in means quality out — a blurry or poorly lit source photo limits what any regeneration can fix. These are strong first-pass images, but for hero shots on major listings you may still want to review them closely before publishing. If you sell in a regulated category, remember that image generation doesn't police your claims — keep any on-image or accompanying copy compliant yourself.

Where this fits

Product photography is the same underlying image workflow used to build AI UGC ads, just aimed at catalog and listing use instead of ad creative. For the mechanics of the Claude-based workflow, see this guide to generating inside Claude via MCP. Review pricing for credit details, or open the app to work from a browser.