Case studies·Tools

Photopea

One developer, zero backend, $50/year hosting

Ivan Kutskir built Photopea alone in the browser — a Photoshop-class editor where the server only serves JavaScript files. Millions of monthly opens, ad revenue in six figures, and hosting cheaper than a nice dinner.

2012–2013

PSD viewer, not a “platform”

Side project · first PSD viewer release (2013)

Kutskir started with a narrow experiment: parse PSD files in the browser and let people toggle layers. No market validation deck, no microservices — just a laptop and curiosity while he was still in university.

He has said he was not counting on money; the project grew because users kept asking for one more feature. Architecture stayed client-side from the first commit.

Lesson

Client-side apps skip an entire category of backend bills. If the hard work is pixels, not CRUD, do not invent a server.

2014–2017

Thousands of hours, still one codebase

2.5M+ visitors · 120k hours used (2017 blog post)

Photopea evolved into a full editor while Kutskir finished his degree — feature by feature, bug by bug. His 2017 “Creating Photopea” post describes solving hundreds of user requests as a solo hobby.

Tiny PHP bits appeared later for optional accounts and PeaDrive cloud storage; core editing still runs entirely in the tab. Most users never touch the server beyond downloading JS.

Lesson

Solo founders win by narrowing scope per release, not by hiring a platform team to orchestrate containers.

2018–today

Millions of opens, hosting costs lunch money

~10M opens/month · ~$100k/month from ads (2021 interview)

In a 2021 Failory interview Kutskir reported Photopea opened about 10 million times a month, with users spending roughly 1.5 million hours in the app — mostly monetized through ads at a few cents per hour of use.

He cited about $50/year for web hosting (static JS delivery) and $16/year for the domain. Revenue later grew beyond that interview, but the stack did not suddenly need Kubernetes — it needed more JavaScript.

Lesson

The cheapest scale move is moving compute to the user’s GPU. Photopea is the ultimate anti-over-engineering flex: Photoshop complexity, static-hosting bill.

Sources

Facts drawn from public engineering posts and interviews. Numbers are approximate where sources disagree — we're stack advisors, not historians.

Get the right stack for YOUR scale — not theirs

Photopea's stack made sense for Photopea. Yours won't — and copying theirs is how you end up with a $4k/month bill and three unused Kubernetes clusters.

Plan my stack — free