Finally launching ArchitectGBT - AI Model Finder - vibe coded it....
Hey 👋
After 3 months of building in public, ArchitectGBT launches (Feb 4) on Product Hunt.
The Story:
I kept hitting the same problem: every AI project started with "which model should I use?" I'd spend 2-3 hours researching, benchmarking, checking compatibility with MCPs, only to pick wrong and start over.
So I built a tool to automate that decision.
What ArchitectGBT Does:
Input your use case (API rate limit sensitive, latency-critical, cost-optimized, etc.) → Instantly get ranked recommendations from 100+ models with MCP compatibility verified.
Export your architecture as code. Deploy to Vercel. Save 3 hours per project.
Tech Stack:
- Next.js + Supabase (obviously)
- Model decision engine (20+ weighted criteria)
- MCP spec crawler (stays updated)
- Vercel deploy API integration
The Business:
Free Tier: 10 credits/month forever + 3 MCP recs/day
Starter Lifetime: $49 (launch price)
Pro Lifetime: $99 (launch price)
Pro Monthly: $15/month (for monthly-only users)
Pro Yearly: $150/year (-17% vs monthly)
Launch window (Feb 4-13): All lifetime deals available + 10 free launch credits per signup.
Try Free Now: architectgbt.com (includes 10 launch credits)
Key Learnings So Far:
- People don't want another "compare models" doc; they want answers
- MCP compatibility is becoming a deal-breaker (good timing)
- One-click deployment is working :)
- Free tier with real credit limits drives conversions better than freemium
What's Next:
- Feb 4: Product Hunt launch
- Feb 11: Email campaign to newsletter (depending on time)
- Feb 15: Twitter/X deep dive thread
- March: Lifetime deal close-out, transition to annual plans
I'm solo-bootstrapped, to discuss business model, pricing strategy, or technical architecture.
Thanks
