Netlify Killed the Seat Tax, Vercel Kept the Credits: Recomputing the 2026 Self-Hosting Breakeven

Search "Netlify vs Vercel 2026" and you'll find the same rule of thumb repeated across a dozen comparison sites: small teams on high-traffic content sites do better on Vercel, larger teams on moderate-traffic apps do better on Netlify, and the two cross over around 800GB of monthly bandwidth on a 5-seat team. It's tidy enough that it spread across JavaScript Twitter as a one-liner — "Vercel taxes your traffic, Netlify taxes your headcount" — and tidy enough that a lot of comparison sites just repeat it without checking the date it was computed. That number is already wrong. Netlify killed per-seat billing entirely on April 14, 2026 — in the same changelog entry that repriced bandwidth — which means the crossover point isn't a fixed 800GB anymore. It's a moving target that scales with how many seats your Vercel team has, and most of the posts citing "800GB" are quoting a rule that predates the update that made team size irrelevant on one side of the comparison.

What Actually Changed on April 14​

Netlify's pricing update did two things in the same release, and most coverage only noticed the first one:

• Unlimited seats. Pro moved from per-seat billing to a flat $20/month for the whole team — Owners, Developers, Publishers, Git Contributors, and Reviewers all included, no incremental cost per person.
• Bandwidth got a lot more expensive per credit. The bandwidth rate went from 10 to 20 credits per GB — doubled — in the same update. Netlify frames this as a wash: the company says it modeled the changes so 98% of customers see a flat or lower bill.

The $20 subscription buys 3,000 credits. At 20 credits/GB, that's 150GB of included bandwidth. Additional credits top up at $10 per 1,500 credits — the same $0.00667-per-credit rate as the included allotment, which means there's no volume discount or overage penalty. Above the 150GB floor, Netlify's bill is just linear: $0.1333 per GB, flat, forever. That's a genuinely different shape than the "$55 per 100GB overage" figure several comparison sites still quote — that number describes an earlier flat-fee overage structure Netlify had already moved away from before this year's update, and it's roughly 4x the current effective rate.

Vercel's model didn't move. Pro is still $20 per seat per month, with 1TB of Fast Data Transfer and a $20 monthly usage credit included per team, and $0.15/GB overage beyond that (the credit typically absorbs bandwidth overage first, extending the effective free tier by another ~133GB). Team size is still the dominant cost driver on Vercel; it's now the only one that matters on Netlify.

The Recomputed Breakeven​

Here's the two formulas, current as of both platforms' official pricing pages:

• Netlify(GB) = max($20, 0.1333 × GB) — independent of team size.
• Vercel(GB, seats) = 20 × seats + 0.15 × max(0, GB − 1,157) — the 1,157GB accounts for the 1TB included plus the ~133GB the $20 credit buys before real overage billing starts.

Setting the two equal and solving for GB (while bandwidth stays under Vercel's 1,157GB free tier, so the crossover is just $20 × seats = 0.1333 × GB) gives a breakeven that scales linearly with seat count — not a fixed number:

The old "800GB on a 5-seat team" rule was close to right for exactly one team size — 5 seats lands at ~750GB, which is why that number kept getting repeated — but it was never a platform-wide constant, and treating it as one means a 3-person team and a 10-person team get the same bad advice. Below its row's threshold, Vercel's flat per-seat base undercuts Netlify's usage-proportional bill; above it, Netlify's $0.1333/GB rate — cheaper per byte than Vercel's $0.15/GB overage — starts pulling ahead. The qualitative version of the old rule of thumb still roughly holds: a small team serving genuinely high-traffic content still tips toward Vercel, a larger team running a portfolio of moderate-traffic apps still tips toward Netlify — it's just that "larger team" now has to mean specifically "more Vercel seats," since Netlify no longer has a seat axis to compare against.

Adding the Self-Hosted Column​

Run the same dozen static/edge apps on a single Hetzner CAX31 instead — 8 ARM vCPUs (Ampere Altra), 16GB RAM, 160GB NVMe, and 20TB of included traffic — and the current price, after Hetzner's June 15, 2026 adjustment, is €20.99/month (about $24.99). That's flat, regardless of team size, and regardless of bandwidth all the way up to 20TB — a ceiling neither hosted platform's realistic static/edge traffic will approach.

Two things fall out of this table. First, self-hosting is competitive with Netlify's $20 floor even at trivial traffic (100GB), and it wins outright everywhere past that — roughly 2.7x cheaper at 500GB, ~5.3x at 1,000GB, and 25-31x by 5,000GB depending on the Vercel team size compared, because 20TB of included traffic keeps the self-hosted line flat while every hosted platform's line keeps climbing. Second, Vercel's per-seat base cost means a small team (3 seats) actually beats Netlify at low-to-moderate traffic even under the new pricing — the seat tax only becomes the losing bet once a team grows past what its bandwidth actually justifies, which is exactly the qualitative split the old rule of thumb was gesturing at.

Put a realistic number on "a dozen static/edge apps" and the table's middle rows stop being hypothetical. A dozen marketing sites, docs sites, and internal tools rarely each pull more than 40-80GB/month individually — a handful of high-traffic launches aside — which puts a typical portfolio somewhere in the 500GB-1,500GB/month range in aggregate. That's precisely the band where Netlify's flat seat elimination and Vercel's per-seat base are fighting for the same customer, and where self-hosting isn't just cheaper on paper — it's cheaper by a factor of 2-10x before any team has to argue about which credit meter is fairest. It's also worth being explicit about scope: this comparison only covers bandwidth and seats, the two axes Netlify's April 2026 update actually touched. Compute, build minutes, and AI inference are separate credit meters on both platforms and aren't priced into any number above — a compute-heavy workload (SSR-rendered apps, background functions) would shift the crossover further in Netlify and Vercel's favor, since a self-hosted box has to run its own compute too, just without a per-invocation meter attached to it.

Per-Seat vs. Credits vs. Flat Is a Shell Game​

Three pricing models, three different tax bases: Vercel taxes headcount, Netlify (post-April) taxes total bandwidth through a credit abstraction, and a self-hosted box taxes neither — you pay for the machine, not for who's on the team or how many bytes it serves this month, right up until you outgrow it. None of these models is "wrong"; they're optimized for different things. Vercel's per-seat charge buys a managed global edge network, first-class preview deployments, and effectively zero operational burden — worth real money to a team that would rather not think about servers at all. Netlify's credit system is genuinely more transparent than a maze of per-resource caps, even if the April repricing made bandwidth pricier per unit. Self-hosting buys neither of those conveniences — you're responsible for TLS renewal, deploy pipelines, and capacity planning yourself, or you're running a platform that handles that for you.

That's the honest tradeoff: self-hosting doesn't win the shell game by picking the best-priced shell, it sidesteps the game by not paying per-seat or per-byte at all. Whether that's the right call depends on whether your team's time is better spent on infrastructure ops or on the 2-3x markup that comes with someone else doing it for you.

Bex.co is the open-source, AI-native Render alternative — push a git repo, get a running HTTPS service on machines you own, at a flat cost that doesn't move when your team or your bandwidth does. Star the repo on GitHub or deploy your first app today.

Sources:

• The end of seats: pricing Netlify for 3 billion builders — Netlify Blog
• Pricing updates for Credit-based plans — Netlify Changelog
• How credits work — Netlify Docs
• Pricing and Plans — Netlify
• Vercel Pro Plan — Vercel Docs
• Pricing on Vercel — Vercel Docs
• Hetzner Price Adjustment 15 June 2026 — Hetzner Docs
• Hetzner Cloud — Cost-Optimized plans