Making bulk product images is limited by review more than by generation. A 200 SKU catalog at four images each is 800 files. The model you pick changes the cost of the run, and every file still needs an eye on it before it reaches a client. Generation takes minutes. Review takes days. Price the job on review time.
Agencies get this backwards on the first big run. The pitch is that AI collapses a catalog shoot into an afternoon, so the quote goes out at afternoon rates. Then someone opens folder three of eleven and finds the model has quietly changed a stitch color on 40 files, and the margin is gone.
This guide covers how to price a catalog run, the throughput lever to check first, how to keep client work separate, and the three specific places bulk generation breaks. If you are still choosing a tool, start with the three kinds of bulk image generator.
Key Takeaways
- Review is the bottleneck. Plan a batch around how many results you can approve in a day. The rest of the schedule follows from that number.
- Pick the model before you quote. The model changes the cost of every image, and every reject costs as much as a keeper.
- Parallel runs set the calendar. Free and Basic run one job at a time, and Ultra runs sixteen.
- Batch only what is identical. A batch is only safe when every item in it takes the same treatment.
- Sample before you commit. Ten SKUs across the range tells you what 800 will look like.
- Keep one brand profile per client. The workflow reads that profile on every run.
Three jobs called bulk product images
Three different jobs get called bulk product images, and each one fails in a different way.
| Job | What it is | Where it breaks |
|---|---|---|
| Volume | The same treatment across many SKUs | Products that need different treatment get it anyway |
| Variation | Many treatments of one SKU | Output drifts from the source product |
| Rescue | Fixing inconsistent inherited images | The source files are not good enough to fix |
Most agency catalog work is volume plus rescue. A client hands over 200 supplier images shot by four different people, and the brief is to make them look like one catalog. That is a standardization problem with a generation step in it. Dropshipping stores face the same supplier-photo problem, covered in one look across every dropshipping SKU.
Naming which job you are doing decides everything downstream, because the review criteria are different. Volume work is checked for consistency. Variation work is checked for fidelity to the product. Rescue work is checked for whether it improved anything at all. A written product photography workflow sets which stages a rule handles and which a person checks.
The cost of a bulk run
In DesignerBox the cost is shown before the run, and the model you pick changes it the most. On one image the difference is small. On 800 it decides which plan you need, so count the images first and budget the rejects with them. Counting the images before a catalog photoshoot shows the sum by products, colors, shots and channels.
| Catalog | Total images | Images to run at a 20% reject rate |
|---|---|---|
| 20 SKUs, 4 each | 80 | 100 |
| 50 SKUs, 4 each | 200 | 250 |
| 100 SKUs, 4 each | 400 | 500 |
| 200 SKUs, 4 each | 800 | 1,000 |
| 500 SKUs, 4 each | 2,000 | 2,500 |
Plans and credits are on the pricing page.
Read that table against your plan’s monthly credits and the model choice is the whole story. Run a small sample on two models, read the cost of each run, and multiply by the right row. Then quote. A large catalog on a premium model can need more than one month of credits, and credit packs add more on a paid plan.
Two things that table hides. Credits reset monthly and do not build into an unlimited pool, so a large catalog is a scheduling problem as well as a budget one. And every failed generation costs the same as a successful one, so at a 20% reject rate you run 25% more images than you keep.
That reject rate is the number worth measuring on your first run. It is the difference between a quotable process and a hopeful one. Against a traditional shoot the comparison still favors this heavily, and what a product photoshoot costs has the day rates to compare against.
Uploading your own photos and the commercial license start on the Pro plan. It is the floor for any client work, whatever the volume.
Parallel generation is the throughput lever
Check this setting before you quote. On a catalog it decides your calendar.
Parallel runs are 1, 1, 4, 8 and 16 across Free, Basic, Pro, Premium and Ultra. A Basic plan processes one image at a time. An Ultra plan processes sixteen.
Run the same 800 image job on both and the queue behaves completely differently. With sixteen runs at a time, the queue clears up to sixteen times sooner than with one.
The practical rule for agencies: the tier decision on a catalog job is a scheduling decision before it is a credit decision. Work out the deadline first, then pick the tier that clears it, then check whether the credits also cover it.
Batch only what is identical
The most expensive mistake in bulk work is batching things that look like they belong together. For photos you already have, the batch edit photos guide covers five ways to edit a set at once.
A batch is safe when every item in it takes exactly the same treatment. Same background, same light, same crop, same framing logic. The moment one item needs a judgment call, it is not part of that batch. Record those treatments as values first. A spec for consistent product images lists the six layers to fix.
Sort a catalog before you run it:
- Group by treatment first. All items that need a plain white packshot go together, whether they are mugs or shoes.
- Pull out the awkward shapes. Anything very tall, very wide, reflective, transparent or fine-detailed goes in a manual pile. Glass and jewelry are the usual offenders.
- Sample ten across the range. Two easy, two awkward, six typical. Run them, review them properly, and only then commit the rest.
- Fix the brief first. If eight of ten samples are wrong the same way, that is a brief problem. Adjust once and rerun rather than correcting eight files.
- Run the full batch in blocks. 50 to 100 at a time, reviewed before the next block starts, so a systematic error costs one block rather than the catalog.
Step 3 is cheap. Ten samples cost a small share of what 800 images cost. When the tool itself is still on trial, an AI proof of concept on about 100 of your own products is the longer version of the same test.
In DesignerBox a saved workflow runs the same way on the next product. Batch runs one workflow over a whole sheet of products, and you keep or discard per row and re-run one row on its own. For inherited supplier images, the product imagery templates are a place to start. Teams driving this from an AI chat such as Claude, ChatGPT or Cursor use the DesignerBox MCP server.
Reviewing 800 images without reviewing 800 images
You cannot skip review, but you can make it cheap.
- Contact sheets first. Review 40 thumbnails on one screen. Consistency errors are visible at thumbnail scale, which is the whole point of the grid.
- Check the product first. Color, texture, logo, proportion, material. Everything else is taste and the client will tell you.
- Spot check at full size. One in ten, plus every awkward item from your manual pile.
- Watch for systematic drift. Errors in bulk work are rarely random. If one is wrong, look for the other 39 that are wrong the same way.
- Keep the source photo beside the result. Fidelity is a side-by-side comparison.
That fourth point is what makes bulk review survivable. Individual errors are expensive to find. Systematic errors are cheap, because finding one finds all of them. Where AI product photo accuracy drifts covers what to look for first, and it is usually fabric and hardware rather than anything obvious.
Keeping clients separate
This is the agency objection, and it is usually phrased as “we already use several AI tools”. The way the work is organized causes most of the trouble.
A stack of single-purpose tools produces client work that lives in whoever’s account generated it. Six months later a client asks for the same campaign on a new product, and the person who built it has left, the prompts are in a Slack thread, and the brand kit is a PDF someone emailed.
What fixes it is organizational: per-client brand profiles that hold the look, workflows that run the campaign again for the next product, and one Assets library for the files. Brand profiles is the mechanic, and running creative for multiple clients without drift covers the operational side.
Seats are worth stating plainly. Team features, shared brand kits and white label are on the Ultra plan, and every plan below Ultra is one seat. For a three person studio that is the real decision, and how agency creative pricing works puts it against retainer math.
If you are already paying for several separate tools, the six tool stack runs that arithmetic.
Where batch generation breaks
Three places, consistently.
The source images are not good enough. Rescue work has a floor. If the supplier photo is soft, small or shot under mixed light, a batch run produces 200 consistently mediocre images instead of 200 inconsistently mediocre ones. Check the worst 10% of the source set before quoting.
The product needs judgment. Reflective, transparent, very fine detail, or anything where the material is the product. These do not batch. Price them separately as manual work rather than discovering them mid-run.
The client has an unwritten rule. Every brand has one. The logo always faces left. The lid is always on. Shadows always fall right. It is never in the brief, it surfaces at review, and it invalidates the batch. Get it out of them during the sample stage, which is the real reason the ten sample step exists.
None of those are fixed by a better model, and it is worth being direct about that before a client expects otherwise. What they are fixed by is sequencing: sample, extract the rules, then run.
For the video equivalent of this problem, where file constraints add a second layer, TikTok Shop product videos and the 10 MB cap shows how a platform limit reshapes the whole production plan. On VTEX the limit is a file size window, and VTEX product images covers it.
Getting started on a real catalog
The shortest path:
- Pull 10 SKUs across the range and run them on the Pro plan, where uploading your own photos starts.
- Review at full size, extract the client’s unwritten rules.
- Fix the brief, then save it as a workflow so the next 190 SKUs run the settled version rather than a fresh judgment call.
- Pick the plan by deadline, using the parallel run numbers. Pick the model by the sample cost.
- Run in blocks of 50 to 100, reviewing between blocks.
Step 3 is the one that compounds. Build the job once against the client’s brand and their own product photos, so nobody re-decides the background halfway through. AI product photography covers the catalog shots themselves, and scaling creative production without adding headcount is the wider version of this problem.
DesignerBox is AI creative production for brands and agencies. Anyone can make an AI picture. Making hundreds that still look like your brand is the hard part. The full workflow from the first product photo to the finished ad, in one subscription. The templates, the workflows, batch, the editors, the brand profiles and the Assets library are all in it. DesignerBox for agencies sets out how an agency runs one workflow per client brand. For an agency, the agency partner program adds partner pricing and onboarding.
FAQ
How many product images can AI generate at once?
Your plan’s parallel run limit sets it more than the model does. DesignerBox runs 1, 1, 4, 8 and 16 runs at once across Free, Basic, Pro, Premium and Ultra. Credits set the total you can produce in a month, and the model you pick changes the cost of each image.
What does it cost to generate 1,000 product images?
The model you pick sets most of it. DesignerBox shows the cost of a run before you start it, so run ten images on the model you plan to use and multiply. Plans and credits are on the pricing page. Pick the model before you quote, and budget for a reject rate on top, because failed generations cost the same as successful ones.
Is bulk AI product photography accurate enough for client work?
For volume and rescue work on straightforward products, yes, provided you start from the client’s own product photography and review the output. It needs review to be reliable, and it is weak on reflective, transparent and fine-detailed products. Sample ten SKUs before committing to a catalog.
How do agencies keep client brands separate?
With a brand profile per client rather than an account per client. The profile holds the look, and the workflow reads it on every run, so the same campaign runs again for the next product. That is what stops the brand living in one person’s prompt history.
How long does a bulk product image run take?
Generation time is set by your parallel limit. Review time is set by your team, and it is the larger number. Measure your own team on the first block and plan from that. Whatever the number turns out to be, an 800 image catalog is a multi-day approval job at any generation speed.
Do I need a commercial license for client work?
Yes. The commercial license on DesignerBox starts on the Pro plan. Produce any images you deliver to a client on Pro or above, however small the run.
Can I automate bulk product images without using the interface?
Yes. DesignerBox has 68 tools over MCP, so an AI chat such as Claude, ChatGPT or Cursor can run the same workflows and apps. This route suits a catalog process that is already stable. The stages a scripted run has to clear, and the platform limit that caps each one, are set out in what actually automates in a catalog run.
Sources
DesignerBox parallel run limits, seat rules and feature gating from the DesignerBox pricing page (designerbox.ai/pricing), September 2026. Plan details change, so confirm on the pricing page before quoting a client. Individual results vary.
