Home About Blog Forum Shop
Sign In Sign Up
Forum Tech & Development What tech stack would you choose if starting a Saa...
G

What tech stack would you choose if starting a SaaS product today in 2025?

Gopi K 1 week ago 243 views 3 replies Tech & Development
I'm planning to build a new SaaS product and I'm doing a fresh evaluation of tech stacks. The landscape has changed a lot in the past 2 years with AI-assisted coding, new frameworks, and the rise of serverless. Here's what I'm considering: **Option A: Laravel + Vue/React (traditional full-stack)** - Proven, mature ecosystem - Great for rapid prototyping - Strong ORM and auth out of the box - Easier to find developers **Option B: Next.js + Supabase (modern serverless)** - Excellent developer experience - Built-in auth, database, storage - Edge functions for global performance - TypeScript-first **Option C: Django + React (Python-first)** - Great if you need ML/AI integration - Strong admin panel out of the box - Large ecosystem **Option D: Go/Rust backend + React frontend** - Maximum performance - Lower hosting costs at scale - Steeper learning curve My use case: B2B SaaS, ~500 initial users, needs good API performance, some AI features, and I want to move fast. What would you choose and why? I'm especially interested in hearing from people who've actually shipped products with these stacks.
tech stack SaaS development React Laravel Next.js
3 Replies
G
Gopi K 1 week ago
Laravel + Livewire or Inertia.js for B2B SaaS under 10k users. You can ship incredibly fast, the ecosystem is mature, and you don't need a separate frontend framework. I've shipped 3 SaaS products with this stack and it's never been the bottleneck.
G
Gopi K 1 week ago
Next.js + Supabase is my recommendation for 2025. The developer experience is excellent, Supabase handles auth/database/storage out of the box, and Vercel deployment is seamless. The main downside is vendor lock-in with Supabase.
G
Gopi K 1 week ago
Don't overthink the stack. The best stack is the one you know best. I've seen people waste months evaluating stacks instead of building. Pick something you're comfortable with and ship. You can always refactor later.