NEW YORK, Sept. 9 — For two days, technology journalist Jack Wallen set out to do something many have promised but few have delivered: build a working web app for a family member without writing a single line of code or spending a dime. His sister is launching a service to help senior women reclaim their beauty.
She needed an app that could offer guidance on clothing types for specific body types, subtle makeup advice, hair styling, posture and communication tips — and it had to be installable on a Linux desktop via AppImage. Wallen, writing in ZDNET, documented his search for an AI tool that didn’t demand a paid subscription.
He found his answer in Dyad, a free, open-source, local AI app builder. According to the project’s GitHub page, Dyad is “a local, open-source AI app builder. It’s fast, private, and fully under your control — like Lovable, v0, or Bolt, but running right on your machine.” For Wallen, that local aspect was the selling point — a privacy advantage over cloud-based rivals.
He installed Dyad on Pop!_OS Linux, though the software also ships for macOS and Windows through installer files on the official download page. Linux users can grab RPM, DEB, or AppImage packages.
Dyad lets users plug into a range of AI model providers, including free services such as OpenRouter, which requires nothing more than a free account and an API key. Wallen chose OpenRouter for his build — and promptly hit the service’s free-tier limits, which paused the project mid-build. He clicked Retry the next day.
The build resumed and completed after about an hour. Wallen tested the resulting app inside Dyad and found it worked exactly as he’d asked. He then published it on GitHub and cloned it successfully.
A personal project, a practical test
The exercise wasn’t abstract for Wallen. His prompt to Dyad was deeply personal, shaped by his sister’s upcoming venture. The tool had to interpret a real-world request — beauty guidance for senior women — and translate it into a functional, installable application.
That it did, without code and without cost. Wallen’s account stands as a practical demonstration that the barrier to building software is lowering fast.
Dyad itself remains free to use. An optional Pro license costs twenty dollars per month and unlocks additional AI models, an advanced agent mode, and auto-debugging.
What to watch next
Wallen’s experiment follows his earlier tinkering with AI tools, including using OpenFactory to build a Linux distro overnight. The pattern is worth noting: local, open-source AI builders are moving from novelty to utility. For families like Wallen’s — where a sibling needs a tool and a brother knows where to look — the question is no longer whether AI can help. It’s how much longer the “no code, no cost” promise will hold as these platforms mature.
