← Study Guide·🏗️ Part VI: Architecture & Business
19

Pricing Model — In Depth

19. Pricing Model — In Depth

Understanding Vercel's pricing is essential for SE conversations. Customers frequently ask about it and compare to alternatives.

19.1 Plan overview

PlanCostWho it's for
HobbyFreePersonal, non-commercial only
Pro$20/user/monthProfessional developers, commercial projects
EnterpriseCustom (~$20–25k/year minimum)Teams needing SSO, SLA, WAF, compliance

19.2 Hobby plan (free)

Strict limits that matter:

  • Function execution: 60 seconds max
  • 100GB Fast Data Transfer/month
  • 1M Edge Requests/month
  • 1M Function invocations/month
  • 4 hours Active CPU/month
  • 360 GB-hours Provisioned Memory/month
  • Commercial use prohibited (violates ToS)

Common trap: Teams build a product on Hobby, launch, go viral, hit limits instantly.

19.3 Pro plan ($20/user/month)

What's included:

  • $20 monthly usage credit (offsets usage charges)
  • 1TB Fast Data Transfer/month
  • 10M Edge Requests/month
  • Function execution: up to 300 seconds (Fluid Compute)
  • Spend management with configurable hard limits
  • Unlimited free viewer seats

Overage rates (examples):

  • Edge Requests: $2/million above included
  • Fast Data Transfer: $0.15/GB above included
  • Active CPU: varies by region (e.g., $0.221/hour in São Paulo)

19.4 What triggers cost growth (the "gotchas")

For customers to watch:

  1. Edge Requests: Every request to the CDN — even for static assets — counts. High-traffic sites with many assets can accumulate quickly.

  2. Active CPU: I/O-bound workloads are cheap with Fluid (pay only for CPU), but CPU-intensive tasks (image processing, heavy computation) accumulate fast.

  3. Fast Data Transfer: Serving large assets (images, video) through Vercel CDN. Guide customers to use CDN-optimised media services (Cloudinary, Mux) for large files.

  4. ISR revalidation: Each ISR revalidation invokes a function. High-traffic sites with short revalidation windows can generate significant function invocations.

  5. AI streaming: Long LLM responses stream for 10–60 seconds. With Fluid Compute, you pay only for CPU — but without it, the cost per AI request can be high.

19.5 Spend management

Vercel provides built-in cost controls:

  • Default $200 on-demand budget with email/SMS/web notifications
  • Configurable hard limit: Vercel auto-pauses all projects when reached
  • Per-project spend tracking in the dashboard

19.6 Pro → Enterprise trigger points

Recommend Enterprise when the customer needs:

  • SAML SSO
  • 99.99% SLA guarantee
  • Managed WAF
  • Audit logs
  • Multi-region compute
  • HIPAA/PCI compliance
  • Dedicated support