Short answer / recommendation
Use a hybrid approach: use Leonardo AI (or a product-focused engine) as your workhorse for bulk white‑background product shots and templated lifestyle composites, and use Midjourney for a smaller set of high-impact branded lifestyle images where artistic flair matters. This gives the best balance of cost, speed, and visual consistency across 500 SKUs.

Why this recommendation (one-sentence rationale)
Leonardo tends to provide features and templates oriented to product realism, reference-based generation, and API/batch flows that reduce per-image manual work; Midjourney excels at creative, editorial-style lifestyle images but usually needs more prompt iteration and manual curation.

Decision criteria (what to weigh)
- Consistency & templates: prioritize engines that support reference images, style-locking, seeds, and project templates. This favors Leonardo for large SKU sets.
- Photorealism for white-background shots: prefer tools with reference matching and controllable lighting; also rely on real product photos for best accuracy.
- Speed & automation: check for an API, CLI, or bulk/batch features and fast queueing—critical for 500 SKUs.
- Cost per final image: include generation, upscaling, retouching, and human QC time.
- Control & editability: ability to inpaint, mask, upload references, and re-run with deterministic seeds.
- Output workflow fit: how easily outputs plug into background removal, color correction, and e‑commerce asset management.

Best-for / Avoid-if
- Best for Leonardo: bulk white-background images, templated lifestyle composites, automated batch pipelines, teams that value speed and consistency.
- Avoid Leonardo if: you need highly stylized or editorial lifestyle art where an artistic, painterly look is central.
- Best for Midjourney: brand-led lifestyle hero shots, unique creative direction, and when you can afford more iteration and manual selection.
- Avoid Midjourney if: you need thousands of near-identical product photos with pixel-perfect consistency.

Practical checklist to run this at scale (500 SKUs)
1. Collect 1–2 high-quality reference photos per SKU (top/side/product detail) and standardized product metadata.
2. Define two templates: white-background product template (lighting, shadow, scale) and lifestyle template (context, props, model look, camera angle).
3. Pilot with 20 SKUs: generate both templates, iterate prompts, lock seeds/styles.
4. Choose engine per template: bulk white shots → Leonardo; hero lifestyle → Midjourney (sample 1–3 variations per SKU).
5. Automate: use API or upload lists to queue generation; use prompt templating (placeholders for color, sku, prop list).
6. Post-process: automatic background-check, remove/replace as needed, color-correct, upscale once.
7. QC batch: sample-check 10% of outputs; flag and regenerate failures.
8. Integrate outputs into your DAM/PIM with naming conventions and metadata.

Budget/skill/team notes
- Small teams / low budget: prioritize the product-focused engine and stricter templating to reduce manual curation.
- High brand/marketing priority: allocate some budget for Midjourney-driven hero images plus professional retouching.
- If you lack prompt-engineering skills, plan for a short pilot and one power user to lock templates.

If you want, I can draft a sample prompt template for both a white-background product shot and a lifestyle variant to pilot across 20 SKUs.

Compare Midjourney and Leonardo AI