AI photos or a photographer? It depends which photo.
The useful comparison isn’t “which is better” — it’s which shots each one is for. Here are the real 2026 numbers, and a straight list of the work you should still hire a human for.
| Path | Cost per photo | Turnaround | Best at |
|---|---|---|---|
| Photographer, per photo | $25–$120 each | Days to weeks | Styling decisions that haven’t been made yet |
| Photographer, day rate | $500–$2,000 a day | Booked in advance | A whole catalogue in one consistent sitting (15–40 finished shots) |
| Per-photo studio service | $30–$60 each | Ship it, then wait | Standardised white-background compliance work |
| DIY with a lightbox | $40–$150 once, then your evenings | Two weekends to get decent | Hard-sided products, if you enjoy it |
| Phoxel | $0.16–$0.30 each | Minutes | The routine shots where the decision is already made |
Photographer and studio figures are the ranges we published in The real cost of product photography in 2026. Phoxel’s range is the arithmetic of its own packs: 10 credits make one photo, so $6 for 200 credits is $0.30 a photo and $49 for 3,000 is $0.16 — see pricing.
Which job is this?
Hire the photographer
These are not edge cases. They are the work a camera and a person do better, and no amount of credits changes that.
Anything with a person in it.
Phoxel does not generate people at all — not a hand holding your product, not a model wearing it, not a bystander in a café scene. If the shot needs a human, you need a human.
The campaign image that sets the brand.
The hero on the homepage, the launch shot, the one that has to feel like a decision rather than a photograph. That’s taste, and taste is what you’re hiring.
A product that doesn’t exist yet.
Generation works from reference photos of a real thing. Pre-production samples, one-off prototypes and anything you can’t photograph yet are outside it.
Texture that has to be felt.
Fabric weave, the grain of leather, how a finish catches light at a specific angle. Macro work with real depth of field is still a lens problem, not a prompt problem.
Anything on-site with people around it.
A working kitchen mid-service, a shop floor, an event. That is reportage — a job about being there at the right moment, which no generator can do for you.
Generate it
The shots where the creative decision was already made, and what’s left is volume.
The ninth angle of the same SKU.
The decision was made on the first shot. The rest is repetition, and repetition is where a day rate stops making sense.
The seasonal restyle.
Same product, new background, twice a year, across forty listings. Re-booking a shoot for that is why most shops simply never do it.
Filling the slots you leave empty.
Amazon gives you seven image slots, Etsy twenty, eBay twenty-four. Most sellers fill four, because photos four through twenty-four never justified a shoot.
Finding out whether it’s worth shooting properly.
Twenty photos for $6 tells you which styling actually converts before you spend $500 committing to it.
Generation replaces the middle, not the top.
A photographer’s day rate buys 15 to 40 finished shots and, more importantly, someone deciding what those shots should look like. That decision is worth paying for once. What it isn’t worth is paying for again every time you need the same product on a different background, or the angle you skipped, or a version that suits December.
That’s the honest shape of it: hire for the decisions, generate the repetition. Most shops discover they were never short of taste — they were short of the twenty routine photos that taste was supposed to be applied to.
The calculator takes your product count, shots per product and a real quote, and returns the cost per usable photo either way. Nothing to sign up for.
Questions worth asking first
Try it on one product before you book anything.
Twenty photos for $6 is cheaper than an hour of anyone’s time, and it answers the question far better than another comparison table can.
Start a project — from $6