9 April 2026  ·  updated 19 August 2026

Hamish is Live

Today we launched Hamish Studio: AI-powered app development on a real, portable stack. It has been a long road to get here, and this post covers why we built it, how it works, and where we go next.

Why we built it

We founded Hamish in 2025 with a simple thesis: AI should make people more capable, not more dependent. The gap between an impressive AI demo and a useful tool is enormous. Founders can generate code in seconds but cannot deploy it, connect it to real data, or get support when something breaks. We set out to close that gap.

How Studio works

Studio is the result. Describe what you want to build, and Hamish builds it: React, Tailwind, production-ready code you own and can export. This is not generated scaffolding or a prototype. Apps built on Hamish connect to real APIs, authenticate real users, persist real data, and deploy to real URLs.

What makes Studio different is the stack underneath. Every app runs on portable, standard technology. There is no proprietary runtime, no vendor lock-in, no hidden costs. You can export your complete project at any time on any paid plan and deploy it anywhere. We believe that is how developer tools should work.

Transparent, token-based pricing

Pricing is transparent and token-based. You see the cost before you pay it. Free accounts include a one-off grant of 15 Build tokens and 8 AI tokens to try it out, with no credit card required. Studio paid plans start at €25 per month (Member), and top-up tokens stay available for as long as your subscription is active. No credit traps, no surprise overages.

Backed by Hamish Production

Studio is backed by Hamish Production: managed infrastructure with real engineering support. When an app outgrows its development environment, Production provides dedicated hosting, monitoring, and incident response from engineers who understand the codebase, because AI built it. Production starts at €399 per month and scales with you.

Our own training pipeline

Behind both products is something we are particularly excited about: our own training pipeline. Customers who choose to share anonymised interaction data from their builds and chats help fine-tune Hamish LLM, a model trained specifically on how people build and ship applications. Sharing is optional on every plan and off by default, and your published apps' user data, secrets and payment details are never used. A domain-specific model outperforms a general-purpose one on domain-specific tasks, at a fraction of the inference cost, and the model improves with every customer who opts in. The full detail, including everything we never train on, is in our AI Training Data Policy.

Your code, your data

All of this runs on managed infrastructure with privacy as a default rather than a premium feature. Your code, your data, your choice.

We are building applied AI that ships.