Next.js vs Webflow vs Framer for a brand site in 2026

Framer is the short campaign. Webflow is the marketing site a team edits every day. Next.js is when the site is the product — heavy photography, an ads stack, an agent, or a file layer that has to live next to the pages. We ship Next.js. That is a bias. The table below is still the honest map.

We turned down a brand site last winter. The marketing lead needed a new landing page every Thursday. Nobody on their side would open a pull request. They did not have a photography problem. They had a Tuesday problem.

We told them to hire a Webflow studio. They looked offended. Then they shipped.

That is the only useful way to compare these three. Not “which is best.” Which refusal you can live with.

Pick the tool for the week after launch, not the week of the pitch.

What is each product actually selling in 2026?

Open the homepages. They are not shy.

Framer calls itself an AI design agent. A professional agent sits on the canvas, generates and refines in place, and they say the change stays editable. They also sell a CMS agent, code-from-an-agent, and connections out to Slack, GitHub, Cursor, Claude, Codex. Their own performance strip on that page: LCP 1.1s, INP 95ms, CLS 0.01, 99.99% uptime. The compare page versus Webflow is blunt: “Framer favors visual speed; Webflow favors structured build depth.”

Webflow calls itself “the agentic web platform for modern businesses.” The line that matters is operational: build without filing a ticket. A visual CMS. Templates so the fifth person does not invent a heading size. They publish customer numbers we did not invent: Lattice’s CMO, Elizabeth Walton Egan, on traffic and conversions; Orangetheory’s CIO, Malcolm Greene — “we’re saving upwards of $6 million a year with Webflow.” Petal — 8 redesigns, zero devs. Dropbox Sign — 67% fewer dev tickets. Verifone — 32 global sites in 10 days.

Next.js is not a website builder. It is a React framework: file-system routing, React Server Components, server actions, streaming, route handlers, image and font and script optimisation aimed at Core Web Vitals. Adobe’s Frame.io team is quoted there on UI that responds within 100ms and animation at 60fps. Another engineer on the same page: they now average 0.09 or lower for Cumulative Layout Shift. There is no “marketing team” tab. There is a repository.

Three products. Three buyers. Treating them as interchangeable is how you buy the wrong invoice. What a marketing website costs is the money version of this. This is the stack version.

How do the official prices compare?

Platform fees are not the studio fee. They still tell you who the tool is for.

Webflow’s May 2026 pricing change folded older CMS/Business plans into Premium. Their Team plan at $2,500 a month is not a typo. It is the product admitting that a marketing org is the customer.

Framer’s live prices display in local currency. We will not invent a USD menu. The student page’s $120 Basic is the only official dollar figure we opened on their side. That is enough to say: the platform is cheap. The studio is not.

Next.js plus Vercel looks cheap until photography optimisation, bandwidth, and seats show up. Image transformations on Vercel Pro start at $0.05 per 1,000 after the included amount. A maison site is a photography bill before it is a framework bill. We learned that on Eternal.

When is Framer the right brand site?

When the site is a campaign that has to feel expensive by Friday.

Framer’s canvas is the closest thing the web has to a designer shipping without a handoff. Their own comparison says they start faster and keep every section editable. Agents edit pages, CMS, and SEO on the real site, not in a side draft. Collaboration is realtime plus branching. Hosting is included.

That is the short-campaign job: a raise, a drop, a seasonal offer, a studio’s own promo, a product that needs to be felt for six weeks. Motion is cheap here in the way that matters — hours on a canvas, not a motion spec and a rebuild.

What it will not be, unless you outgrow it carefully: the system a ten-person marketing team lives in for three years. Framer Pro’s public limits (on the pricing page we opened) cap site pages, CMS collections, and items, then sell overages. Advanced hosting, A/B testing, and extra locales are add-ons. Fine for a brand site with a journal. Tight for a localisation machine.

It will also not hand your next engineer a codebase they want. You can put custom code on the canvas. You do not leave with Next.js. Framer’s compare page is honest about this: “Webflow goes deeper; Framer stays easier for site teams.”

Use Framer when the cost of being late is higher than the cost of leaving later.

When is Webflow the right brand site?

When a marketing team has to edit the site the way they edit a doc.

Webflow’s homepage is a list of things marketers were tired of asking engineering for. New pages on their timeline. A visual CMS. Templates so the fifth person does not invent a new heading size. Experiments and personalisation as native tools. AEO — answer-engine optimisation — as a product line, not a blog post.

The quotes they publish are about throughput, not taste. Petal redesigned eight times without developers. Dropbox Sign cut dev tickets by 67%. 21.co cites $300k a year saved. Spin Master cites $500k in reduced dev cost. Orangetheory’s $6 million is the outlier, and it is their CIO speaking, not a designer.

If those sentences sound like your Monday, hire for Webflow. Not because Webflow sites are prettier. Because the organisation is the constraint. A Next.js site with a headless CMS can reach the same place. It will cost engineering every time someone wants a new component. Webflow’s bet is that most brand sites are a component library plus a lot of pages.

What Webflow will refuse: being the cheapest canvas for a one-off (that is Framer), and being the right home for a site that is really an application — custom checkout, an agent runtime, twenty ad variants on one measurement layer you own in git. Code components and Webflow Cloud can force some of that. You will feel the grain.

Their Team plan at $2,500 a month plus Optimize at $299 a month is the honest enterprise shape. If that shocks you, you may be a Premium site at $25 a month and a freelance designer. Valid brand site. Not a platform transformation.

When is Next.js the right brand site?

When the site is the product.

Heavy photography that has to hold on a phone. Type that cannot be a Webflow combo class. A landing page the campaign and the agent both use. A file stack — llms.txt, a real /pricing page, answer-ready copy — sitting next to the routes. An origin you control when the ads pixel, the CRM, and the agent all need to agree.

Next.js image documentation is the photography argument in one component: correct sizes, modern formats, no layout jump, lazy-load off-screen, remote images with allow-lists. That is why Eternal is Next.js. A maison site is a sequence of rooms. The collection is the light. Nothing moves unless it should. A canvas would have let us ship faster. It would not have let us be that exact.

Route handlers and server actions are the ads-and-agent argument. The page, the form, the event, the handoff — one repository. Webflow and Framer will take a form and a script tag. They will not be the system.

What Next.js will refuse: a marketer shipping a new composition on Tuesday without a designer or a developer. Same-day visual play. The cheapest test of a six-week offer. A CMS unless you add one (Sanity, a headless tool, or a deliberately small file-based journal — this site is the last of those).

We ship Next.js. We like repositories. We like the first impression matching the work behind it. That is a taste, and it is a service, and it is not the right taste for every brand site.

If nobody in the company will open a pull request, do not buy a pull request.

How should you choose between Next.js, Webflow, and Framer?

Ask four questions. Then stop arguing about fonts.

Who edits on Tuesday? A marketer who will not wait: Webflow. A designer who wants the canvas: Framer. An engineer who wants review: Next.js.

How long must this site live? Six weeks: Framer. Three years of weekly pages: Webflow. Three years of being the thing people remember: Next.js, if you will staff it.

What is the heavy object? Motion and a feeling: Framer. A CMS and a localisation plan: Webflow. Photography, performance, and a stack that is not a page: Next.js.

Where does the buyer start? Often not on your homepage. That journey is buyers start in ChatGPT. All three tools can publish readable pages. Only a site you own end-to-end makes it easy to ship the files a model actually quotes. That is a Next.js convenience, not a Next.js monopoly — Webflow now sells well-known files on Premium; Framer sells static files on Pro. The difference is how much of the surrounding system lives with the page.

What does a fair comparison table look like?

Bias declared. Rows we would still defend in someone else’s workshop.

Framer’s compare table says they are easier to forecast. Webflow’s pricing page says Team is $2,500 a month. Next.js is free and Vercel is usage. Forecast the organisation, not the SKU.

What do we refuse, because we ship Next.js?

A few things, so the bias is on the page.

We refuse to tell a content-heavy marketing team that Next.js will make them faster than Webflow. It will not, unless they hire the people Webflow is designed to un-hire.

We refuse to rebuild a six-week campaign in a custom app so that we can say we used the framework we like.

We refuse a “headless Webflow” or “Framer plus Next.js” sandwich unless there is a real split: campaign on the canvas, product on the origin. That hybrid is real. It is also two invoices.

We refuse to pretend hosting is the decision. Vercel Pro is $20 per user. Webflow Premium is $25 a month. Framer Basic is $120 a year. The decision is who touches the page, and what the page has to carry.

The web work we take is the last column. Brand and marketing sites when they have to hold. Product surfaces. Next.js, designed and engineered here. Performance that holds up on a phone.

FAQ

Is Next.js better than Webflow for a marketing website?

Not if marketers must ship without a developer. Webflow is built for that week — their homepage is “build without filing a ticket,” and customers like Petal report eight redesigns with zero devs. Next.js is better when the site is the product: photography, custom behaviour, an ads and agent stack, a file layer you own. We ship Next.js. For a daily-edit marketing site, Webflow is often the fair pick.

Is Framer just a cheaper Webflow?

No. Framer is faster to a designed page and weaker as a years-long marketing OS. Their own compare page says visual speed versus structured depth. Official Framer pricing is a site plan plus seats and add-ons. Webflow goes from $15–$25 a month into a $2,500 Team plan. If you need governance, localisation, and a marketing org, that jump is the product. If you need a campaign on Friday, it is overhead.

Can I start on Framer and move to Next.js later?

Yes, as a rewrite, not a export. Plan the move when the site outgrows the canvas: too many integrations, photography that needs a real image pipeline, or an agent that has to live on the origin. Budget a second project. The cheap version is to pick correctly the first time using the Tuesday-edit question.

Does Next.js rank better than Webflow or Framer?

Not because of the logo. Google’s AI features guide still says write for people; no special markup is required. All three can be fast, structured, and readable. Thin pages lose on every stack. Rank and citation come from the words, the files, and the proof, not from React.

What about WordPress?

Still the default for a large share of the web, and not the comparison this page is for. If your constraint is a publisher’s workflow and a plugin ecosystem, WordPress can be the honest tool. If your constraint is a campaign, a marketing team, or a site that is the product, the three above are the 2026 shortlist.

How much more does a Next.js brand site cost?

The framework is free. The studio is not. Tenet’s 2026 custom bands ($15,000–$80,000, enterprise above) and Innowise’s rebuild tiers ($30,000–$80,000, then $100,000–$200,000+) apply to serious Webflow work too. Next.js costs more when you buy engineering the canvas would have hidden. It costs less over three years if you would have been fighting the canvas. Read the cost piece as ranges, not as our menu.

Should the ads landing page be on the same stack as the brand site?

Yes, if you want one measurement story. A Framer campaign next to a Next.js origin is a valid split. Two pixels and two forms is how inquiries disappear. We would rather the landing page share an origin with the brand site and the agent. If marketing must iterate the landing page daily, Webflow or Framer for that surface, with events you can still defend.

If you already know who edits

Send the homepage, who edits it, and what has to live on it besides pages. Write to the studio. We will write back with the stack we would pick, and the one we would refuse.

If the honest answer is Webflow or Framer, we will say that. If it is Next.js, Eternal is the kind of site we mean — not a template with a framework badge.