How to batch-generate product shots with Leonardo AI

Recommendation
Use Leonardo AI as your generator and automate batch runs with a small script or Zapier/Make that feeds a CSV of prompt templates + seeds. Keep a single reusable prompt template with placeholders for SKU, angle, background, and a fixed image size (Shopify-friendly). Use deterministic seeds so each SKU variant is reproducible, and use img2img/inpainting for tight product consistency when needed.

Why this works (short)
Leonardo gives good price/control for image generation. Deterministic seeds + structured prompts let you produce consistent product identity across 100 SKUs while swapping angles and backgrounds. Automation removes manual repetition.

Decision criteria (pick based on your situation)
- Budget: If you need high throughput and reproducibility, pay for Leonardo’s higher tier or API access. If you’re low-budget, do smaller batches manually.
- Skill level: If you have a developer, write a Python script calling the API. Non-devs can use Zapier/Make + CSV and upload images via the web UI.
- Output quality: For pixel-perfect product shots, plan for img2img/inpainting and manual QA; if “good enough” is acceptable, pure text-to-image is faster.
- Team size / workflow stage: Small teams can batch 20–50 at a time. Larger catalogs (100+) should build an automated pipeline and QA queue.

Practical checklist (step-by-step)
1) Define final image specs: Shopify recommends 2048x2048 for square product images (good for zoom). Also export medium/resized versions (e.g., 1024x1024, 800x800) for thumbnails.
2) Create a master prompt template. Example: "{SKU} — close-up of product, {angle}, studio lighting, white seamless background {background_style}, realistic texture: {material_notes}. Camera: 50mm, high detail, neutral color balance."
3) Create a negative prompt block: blurry, watermark, hands, extra limbs, low-resolution, text overlays.
4) Decide seed strategy: choose a base seed for the product family and then offset by variant (seed = base + variant_index). For consistent product identity across angles/backgrounds keep the same base seed and only change the angle/background placeholders; change seed when you want a visibly different render.
5) Make a CSV with columns: SKU, prompt_filled, seed, angle, background, output_filename.
6) Automate: use Leonardo AI’s API or a web automation tool. Script (pseudo): loop CSV rows -> call generation endpoint with prompt + seed + size -> save image with filename.
7) Use img2img/inpainting for edits: upload a high-quality reference mask to maintain exact product contours (good when you need consistent shape across angles/backgrounds).
8) QA pass: sample 10% for color, shadow, reflection consistency; fix via inpainting or re-run with refined prompt/seed.
9) Export/resizing + naming: output master at 2048x2048, generate resized copies and name like SKU_angle_bg_2048.jpg.
10) Upload to Shopify in bulk via CSV import or API.

Best-for / Avoid-if
- Best-for: sellers who need many consistent, photoreal product images and have some automation capability.
- Avoid-if: you need exact physical-dimension accuracy for engineering drawings (use studio photos instead).

Extra tips
- Keep a prompt & seed log so you can reproduce or re-run items. Use ChatGPT to generate prompt variations and to expand your template CSV programmatically.
- Start with a 10–20 SKU pilot to refine prompts and seeds before scaling to 100+.

CTA
If you want to run the pipeline on Leonardo AI, start by creating the prompt template and CSV, then use Leonardo AI’s batch/API features to automate generation. Good luck!

Compare Leonardo AI and Midjourney