Websites and hosting decision

Vercel

A competent developer can reproduce a basic deploy+CDN+serverless workflow; the full Vercel experience (fluid compute, AI gateway, security add-ons, global orchestration and polish) is costly to replicate, so keep paying if you need those features.

Visit website
You pay

$20/mo

$240/yr

Per seat. Read off the official pricing page.

You’d pay instead

$50one-off30 h to build

$25/mo6 h/mo upkeep

On cash alone, building overtakes the subscription at 2 seats.

Open-source builds that already do this

Every project below is open source and already does this job today. Fork one, self-host it, or take the parts you need - the build prompt further down assumes an empty file, and this is the shortcut past that. Licences differ; check the one on each card before you ship. All Vercel alternatives, with the arithmetic →

What a replacement has to do

  • Provide continuous deploys from a Git repo to a global CDN with serverless function execution and basic observability (deploy, preview, rollback).

What it still won’t have

  • Fluid Compute / Active-CPU billing and integrated serverless platform
  • AI Gateway, Agent Stack, and Sandbox features
  • Built-in global deployment orchestration and tenant isolation
  • Managed WAF, Bot Management, BotID and advanced security add-ons
  • Integrated preview environments, toolbar comments, and one-click rollbacks with Vercel polish

What remains hard

  • Infrastructure at scaleShip apps that scale from zero to millions instantly
  • Execution qualityVercel handles infrastructure automatically and provides these tools to help you deploy and scale:
Read the build prompt

First-year cost

Keep paying

Paying is—cheaper in year one.

On cash alone, building overtakes the subscription at 2 seats.

Paid seatsseats

Money you would actually spend

Keep paying
—

Subscription price × seats × 12

Build it
—

AI build —APIs + hosting —

Time you would spend

—

—

What you would spend

What we assumed

The verdict above measures whether you could build it. This one is only about money.

Runnable build prompt

Not run yet
Build a minimal self-hosted Vercel-like deploy platform using: GitHub Actions, Node.js for build hooks, AWS S3 + CloudFront for static assets, AWS Lambda + API Gateway for serverless functions, Postgres (or SQLite) for deployment metadata, and CloudWatch/ELK for logs. In scope: Git->CI pipeline, artifact publish to S3, CloudFront distribution with invalidation, Lambda endpoints wired to routes, a small web dashboard to show current deployment, preview URLs for pull requests, automatic rollback to last successful deploy, and basic request logging and health checks. Out of scope: agentic AI features (AI Gateway, Sandbox, Agent Stack), advanced WAF/Bot management, multi-tenant enterprise features, and deep observability dashboards. Include input validation, retries, error handling for builds/deploys, and unit/integration tests for CI hooks, upload, and function invocation paths.
How we checked4 sources · 3/3 runs agreed · evidence score 64

How the score was reached

  • Partly verdict base52
  • An open-source build was found+5
  • 4 cited sources+3
  • Price verified on pricing page+3
  • 3/3 assessment runs agreed+4
  • Hard moats found in the evidence-3
  • Evidence score64

The base comes from the verdict. Everything under it is a check that either happened or did not, and each one is a fact frozen in this record rather than a judgement made at render time - so the same evidence always produces the same number.

How scoring works →

Cited sources · 4

Every page the run actually retrieved.

Integrity checks

What held up, and what did not.

✓ Price read off the page✓ 3 independent runs, one answer✓ Citations limited to fetched pages! 2 moats quoted from the page