Used For

• +**Your React boilerplate takes 3 hours to set up with auth and database** → Chat "build a Kanban board" and get a working app in minutes
• +**You're copy-pasting code between 4 different AI chat windows** → One conversation handles frontend, backend, and database all together
• +**Cloud AI builders timeout or hit rate limits mid-flow** → Runs on your machine, no network delays killing your momentum
• +Exports clean code to VS Code - no vendor lock-in or subscription required to access your work
• +Works with any AI model through your own API keys - switch from GPT to Claude to local models

Best For

• >Side project stuck at 60% done - boilerplate code after work is soul-crushing
• >Your startup needs 5 MVP prototypes this month and you're one person
• >Tired of v0/Bolt lagging every edit while you lose your train of thought

Not For

• -Teams over 20 people expecting enterprise features like branching, code reviews, or advanced security out of the box
• -Anyone wanting drag-and-drop visual building - this is chat-based only
• -Non-technical users who don't want to deal with API keys or basic setup

Pairs With

• *Supabase (handles your database, auth, and backend functions so you don't need to set up servers)
• *Vercel (where you deploy the finished app because local development doesn't help users)
• *VS Code (where you edit the exported code when you need more control than chat allows)
• *Figma (for designs you want to recreate, though import feature is still coming)
• *GitHub (to version control the code since Dyad doesn't handle git workflows)

The Catch

• !Free tier gives you 50 big changes per day on Gemini Pro - sounds like a lot until you're iterating on a complex feature
• !Large projects eat through AI tokens fast without the Pro version's Smart Context feature
• !You'll still need to add production hardening like proper error handling and security reviews yourself

Bottom Line

Finally, an AI coder that runs locally so you're not waiting 30 seconds for every tiny change.