TL;DR / Key Takeaways
- Hobby is $0 but limited to personal, non-commercial use; Pro is $20/month, which includes 1 deploying seat and $20 of monthly usage credit; extra deploying seats are $20 each.
- Pro includes 1 TB of Fast Data Transfer; past the credit, usage bills on demand, for example $0.60 per 1M function invocations and $0.128 per Active CPU hour in Washington D.C. (iad1).
- A small Next.js app with about 150,000 function calls a month uses roughly $1.72 of metered usage in our worked example, so the bill stays at $20.
- The usual surprise drivers, per Vercel's docs, are function memory billed during I/O waits, image variants, extra seats and add-ons.
- Spend Management (Pro) can pause production deployments at a budget you set; checks run every few minutes, so set it below your true maximum.
Short answer: Vercel has three plans. Hobby is free for personal, non-commercial projects. Pro costs $20 a month for one deploying seat and includes $20 of usage credit, plus 1 TB of data transfer; extra deploying seats are $20 each. Enterprise is custom. For a small Next.js app the $20 credit usually covers all metered usage, so most solo Pro bills are exactly $20; bills grow with seats, add-ons and function-heavy workloads.
How we checked: we read vercel.com/pricing and Vercel's docs on pricing, the Pro plan, the Hobby plan, function pricing, regional pricing for iad1 and Spend Management, then worked the arithmetic ourselves. Prices checked on each vendor's own pricing page on November 1, 2026. US prices exclude tax and many rates vary by region. Some links are affiliate links: if you sign up, the vendor may pay us a commission at no cost to you. It never changes our picks.
Vercel pricing plans at a glance
| Plan | Price | Who it is for | Included usage | Past the included amount |
|---|---|---|---|---|
| Hobby | $0 | Personal, non-commercial projects; 1 developer seat | 100 GB Fast Data Transfer, 1M CDN requests, 1M function invocations, 4 Active CPU hours, 360 GB-hrs memory, 5K image transformations per month | No paid overage; Vercel says you usually wait 30 days for the feature to reset |
| Pro | $20/month (1 deploying seat); extra deploying seats $20/month; viewer seats free | Commercial sites, freelancers, teams | $20 monthly usage credit, plus 1 TB Fast Data Transfer via the included Flat Rate CDN tier | On-demand per-unit rates (table below) |
| Enterprise | Custom | Large companies needing contracts, SSO, SLAs | Custom | Custom |
One discrepancy to know about: the pricing page lists Pro with "10M / month included" CDN requests, while Vercel's Flat Rate CDN docs describe the tier included with Pro as 1M CDN requests and 1 TB of data transfer. If you expect more than 1M requests a month, check the Usage page in your dashboard rather than relying on either number.
Vercel Hobby pricing: what free includes
Hobby costs nothing. The Hobby docs list up to 200 projects, 100 deployments a day, 50 domains per project, 2-vCPU builds with 8 GB memory, function duration up to 300 seconds and 1 hour of runtime logs. There is no billing cycle: if you exceed a limit, Vercel says that in most cases you wait until 30 days have passed before you can use that feature again.
Vercel Pro pricing: the $20 platform fee and credit
Pro is a $20 monthly platform fee that includes one deploying seat and $20 of usage credit. Per the Pro plan docs, the credit applies to all managed infrastructure resources from the first unit, expires at the end of the month and does not roll over. Additional Owner or Member seats are $20 a month each; Viewer seats are free and unlimited. Pro also raises limits (unlimited projects, 6,000 deployments a day, function duration configurable up to 800 seconds) and adds email support, Spend Management and one free first-year domain on eligible TLDs.
Once the credit is used, everything is billed per unit. These are the Pro on-demand rates; Vercel lists most as "starting at" because they vary by region. The figures below are the Washington D.C. (iad1) rates.
| Resource | Hobby included | Pro on-demand rate (iad1) |
|---|---|---|
| Fast Data Transfer | 100 GB/month | 1 TB included; $0.15 per GB on demand |
| CDN requests | 1M/month | Included in Flat Rate CDN tier; $2.00 per 1M on demand |
| Function invocations | 1M/month | $0.60 per 1M |
| Fluid Active CPU | 4 hours/month | $0.128 per hour |
| Fluid Provisioned Memory | 360 GB-hrs/month | $0.0106 per GB-hr |
| ISR reads / writes | 1M / 200K per month | $0.40 per 1M / $4.00 per 1M |
| Image transformations | 5K/month | $0.05 per 1K |
| Image cache reads / writes | 300K / 100K per month | $0.40 per 1M / $4.00 per 1M |
| Fast Origin Transfer | 10 GB/month | $0.06 per GB |
| Build minutes (Basic, 2 vCPU) | Included | $0.007 per minute ($0.0035 per CPU minute) |
| Blob storage | 1 GB/month | $0.023 per GB |
| Web Analytics events | 50K/month | $3 per 100K events (pricing page) |
Fixed-price Pro add-ons, which the credit does not cover: Password Protection $20 per project per month, Static IPs $100 per project per month, SAML SSO $300 a month, HIPAA BAA $350 a month, Preview Deployment Suffix $100 a month, Web Analytics Plus $10 a month and Speed Insights Plus $10 per project per month.
For high traffic, the Flat Rate CDN tiers replace per-GB billing with a fixed fee: included tier 1M requests and 1 TB, then $20/month for 10M requests and 50 TB, $100/month for 50M and $300/month for 150M. Vercel moves you up a tier at the start of the next cycle if a whole month exceeds capacity; a one-day spike does not trigger it.
Vercel Enterprise pricing
Enterprise is quote-only. The pricing page lists every metric as custom. Spend Management on Enterprise requires the Flexible Commitment plan, per Vercel's docs.
Worked example: what a small Next.js app costs on Vercel Pro
Assumptions, all ours: one developer, functions in iad1 with 2 GB of memory, 300,000 CDN requests and 60 GB of transfer a month, 150,000 server-rendered or API function calls each using 40 ms of CPU and keeping the instance busy for 300 ms, 2,000 image transformations and 200,000 image cache reads, 100,000 ISR reads and 20,000 ISR writes (one unit each), and 60 builds of 2 minutes on Basic machines.
| Line | Arithmetic | Cost |
|---|---|---|
| Invocations | 150,000 / 1M x $0.60 | $0.09 |
| Active CPU | 150,000 x 0.04 s = 6,000 s = 1.67 h x $0.128 | $0.21 |
| Provisioned memory | 150,000 x 0.3 s x 2 GB = 25 GB-hr x $0.0106 | $0.27 |
| Image transformations | 2,000 / 1K x $0.05 | $0.10 |
| Image cache reads + writes | 200,000 / 1M x $0.40 + 2,000 / 1M x $4 | $0.09 |
| ISR reads + writes | 100,000 / 1M x $0.40 + 20,000 / 1M x $4 | $0.12 |
| Builds | 60 x 2 min x $0.007 | $0.84 |
| CDN requests and 60 GB transfer | Inside the included tier | $0.00 |
| Metered total | $1.72, covered by the $20 credit | |
| Bill | $20 platform fee | $20.00 |
At ten times the traffic (1.5M calls, 20,000 transformations, 200,000 ISR writes, 3M requests, 600 GB) the same lines add up to about $9.60, still under the credit. The caveat is CDN requests: 3M is above the 1M tier in the Flat Rate CDN docs, which would move you to the $20 tier the next cycle (or $2 per extra 1M on demand), but inside the 10M the pricing page lists. A second developer adds $20 either way.
Where a bill really moves is compute that waits. Vercel's function pricing docs say Active CPU pauses during I/O but provisioned memory keeps billing. An AI feature making 1.5M calls a month at 500 ms CPU and 3 seconds of instance time each (mostly waiting on a model) works out to 208 CPU hours ($26.67) plus 2,500 GB-hrs ($26.50) plus $0.90 of invocations: about $54 of usage, so roughly $54 a month instead of $20. If you are building with AI coding tools, our post AI coding's hidden tax: my $800 Vercel mistake covers how that happened to us.
Enjoying this? Get one like it in your inbox each morning.
one email a day · unsubscribe in two clicks · no third-party tracking
What usually causes a surprise Vercel bill
- Memory during I/O waits: Vercel bills provisioned memory for the whole instance lifetime, including time spent waiting on databases or AI APIs.
- Image variants: Vercel's image cost guide says each distinct cache key creates its own transformation, short cache TTLs re-bill in every region, and unused transformations are wasted spend. Its fixes include a longer minimumCacheTTL, one format, one quality and fewer deviceSizes.
- Seats: every Owner or Member seat beyond the first is $20 a month; only viewers are free.
- Add-ons: Password Protection, Static IPs, SSO and similar are flat monthly fees and are not counted by Spend Management.
- Regions: rates vary by region; Active CPU is $0.128 per hour in iad1 but $0.221 in Sao Paulo (gru1), per Vercel's function pricing table.
How to cap spend on Vercel
- Open Settings, then Billing, then Spend Management (Owner or Billing role, Pro or Enterprise on Flexible Commitment).
- Set an On-Demand Budget. New Pro customers get notifications at $200 by default; alerts go out at 50%, 75% and 100%, with SMS at 100%.
- Turn on Pause Production Deployments. At the budget, production sites return a 503 until you resume each project by hand.
- Or use the CLI: vercel budgets set --amount 200 --pause (vercel@59.23.0 or later).
- Set the budget below your real limit. Vercel checks spend every few minutes, and pausing does not stop storage charges, AI Gateway or v0 usage.
On Hobby there is nothing to cap: you cannot be charged, and features stop when limits are hit.
Is Vercel worth it, or should you look elsewhere?
Pick Vercel Pro if you run a commercial Next.js app and work alone or with one or two developers: the $20 credit and 1 TB of transfer cover a lot, and Next.js is maintained by Vercel. Stay on Hobby for genuinely personal projects.
Look at Netlify if you have a larger team, because Netlify Pro is $20 a month with unlimited members, versus $20 per deploying seat on Vercel. Netlify meters bandwidth from its credits, so it can cost more than Vercel for heavy traffic; our Vercel vs Netlify comparison runs both at the same usage. For long-running servers, Railway bills per second of CPU and memory. More options are in our Vercel and Netlify alternatives guide, and Netlify's plans are summarized on its Stork page.
Frequently asked questions
How much does Vercel cost per month?
Hobby is free. Pro is $20 a month, which includes one deploying seat and $20 of usage credit; each additional deploying seat is $20 a month and viewer seats are free. Enterprise is custom-priced. Usage beyond the Pro credit is billed on demand at the rates on vercel.com/pricing.
Is Vercel free for commercial use?
No. Vercel's fair use guidelines say Hobby teams are restricted to non-commercial personal use and that all commercial usage requires Pro or Enterprise. Their examples of commercial use include taking payments, advertising a product or service, being paid to build or host the site, sites whose main purpose is affiliate linking, and showing ads. Asking for donations is not commercial use.
What does the Vercel Pro $20 credit cover?
Per Vercel's Pro plan docs, the $20 monthly credit applies to all managed infrastructure resources from the first unit, such as Active CPU, provisioned memory, invocations, image optimization and ISR. It expires at the end of each month and does not roll over. Seats and add-ons are billed separately.
What happens if I go over the Hobby limits?
Vercel's Hobby docs say that in most cases, if you exceed a usage limit, you have to wait until 30 days have passed before you can use that feature again. There is no option to pay for extra usage on Hobby; you upgrade to Pro for that.
Can I set a hard spending cap on Vercel?
Yes, on Pro. Spend Management lets you set an On-Demand Budget and turn on Pause Production Deployments, which stops your production sites serving traffic once the budget is reached. Vercel checks spend every few minutes, so pausing can lag, and storage, seats and add-ons keep billing.
How much does Vercel charge for bandwidth?
Hobby includes 100 GB of Fast Data Transfer a month. Pro includes 1 TB; Vercel's pricing page lists on-demand Fast Data Transfer starting at $0.15 per GB, and its Flat Rate CDN docs list paid capacity tiers from $20 a month for 10M requests and 50 TB.