Dyad
Free, local, open-source AI app builder — like Lovable or v0, but running on your own machine.
About
Dyad is an Electron desktop app that builds full-stack applications from natural-language prompts, entirely on your local machine. You bring your own model API keys (OpenAI, Anthropic, Gemini, Ollama, LM Studio are all supported), so there is no vendor lock-in and your code never has to leave your computer. It launched in April 2025 by solo founder Will Chen, an ex-Googler, and picked up more than 20,000 GitHub stars within months after a well-received Show HN.
The honest picture: the core repo is Apache 2.0, but code under src/pro is under the Functional Source License, so not everything is fully open. As a local Electron app it is heavier than a browser tab, and generated apps still need a real developer to review them — Dyad is a fast starting point, not a replacement for engineering judgment. The project is young and its long-term maintenance depends on a small team.
What it can do
- Prompt-to-app generation running fully locally
- Bring-your-own-keys: OpenAI, Anthropic, Gemini, Ollama, LM Studio
- Exports clean, runnable code you own
- Visual iteration on generated apps
- No account or cloud dependency required
Setup
Download the installer for Mac or Windows from dyad.sh — no sign-up required. Add your own AI provider API keys in settings, describe the app you want, and Dyad scaffolds and iterates on it locally.
Prerequisites
A Mac or Windows machine; your own API keys for a supported model provider (or a local model via Ollama/LM Studio).
Use cases
Works with
Trust score — 8/10
Scored 8/10 (verified tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 52; Trending score 85/100 from researcher assessment.
★ 21,500 GitHub stars
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.