Webflow vs Next.js: when should a B2B site move to code?

Stay on Webflow while your marketing team needs to ship pages without an engineer and the site's job is marketing. Move to Next.js, usually with a headless CMS such as Sanity, when the site becomes part of the product, needs logic Webflow can't run, outgrows Webflow's CMS and locale limits or its seat costs, or when your engineers already own a Next.js codebase. If only one piece needs code, such as a customer portal or a calculator backed by your data, split it: Webflow for marketing pages, Next.js for the app on a subdomain or a path.

Prices and limits below are as of October 2026, linked to the vendors' own pages. For the wider field, see our 2026 website platform comparison.

Our bias, stated once: AGR Studio builds on Webflow, Next.js, Framer, WordPress and Astro, we're an Official Framer Expert, and our own site runs on Astro + Sanity.

How do Webflow and Next.js compare for a B2B site?

Webflow is a hosted visual builder with a CMS. Next.js is an open-source React framework (MIT, maintained by Vercel): you bring the CMS and the hosting. The current major is Next.js 16, released 21 October 2025; 16.4.3 is the latest stable release as of October 2026 (Next.js blog).

When should a B2B team stay on Webflow?

Stay while the site is a marketing site and the people who change it are marketers. That's what Webflow is for: a designer builds components once, and the marketing team assembles and edits pages without a ticket to engineering. When our clients picked a visual builder (usually Framer), the reasons they gave apply to Webflow too: the team edits without a developer, and hosting is included.

Webflow also reaches into code now. Premium includes code components, React components imported into the Designer, which Webflow server-renders by default. One interactive widget is not a reason to rebuild.

Our rule: if next year's roadmap is mostly landing pages, copy changes and blog posts, stay and spend the money on content. That's the kind of site we build as a Webflow design agency; headless CMS vs website builder goes deeper on the trade.

What are the signs a team has outgrown Webflow?

One signal is a nuisance. Two or three that keep coming back make a migration case.

You're close to the CMS or locale ceiling

Webflow's Premium plan holds 20,000 CMS items in 40 collections. The Team plan, at $2,500 a month on an annual contract, raises collections to 100 but still lists 20,000 items, so a growing content library has no cheap next step.

Languages are priced per locale: Localize Essential is $9 per locale a month billed yearly, for up to 3 locales; Advanced is $29 per locale for up to 10, with localized URLs and visitor routing (pricing). Six languages on Advanced cost $174 a month. Sanity has quotas too (Growth includes 25,000 documents), but it lists "unlimited locales" on every plan (within attribute limits), so language count stops being a line item.

New editors cost a seat since October 2026

As of 5 October 2026, Webflow's legacy Editor is fully deprecated: anyone who edits content needs a seat on the site's Workspace. Editors moved over from the old Editor got free seats, and sites kept in an agency's Workspace get free client seats; a new editor on your own Workspace is a paid seat. A limited seat (edit content, or build pages from components) is $15 a month billed yearly or $19 monthly; a full seat is $39 or $45 (pricing). Five new content editors add $75 to $95 a month.

To be fair: Sanity Growth is also $15 per seat, with free Viewer-role users (Sanity pricing). The seat change is a reason to re-examine the platform, not a savings case on its own.

You need app features

This is the strongest signal. Logins, gated content, per-account pricing, personalisation by company, calculators backed by your own data: these need code that runs on a server, and Webflow's Code Embed takes HTML, CSS and JavaScript but "can't run server-side languages". Webflow also retired its own gated-content feature, User Accounts, on 29 January 2026, and points customers to third-party tools such as Outseta and Memberstack. That covers a members' area. When the logged-in part is your product, it belongs in your product's codebase.

In Next.js these are ordinary features: pages are Server Components by default, so they can fetch from databases or APIs on the server, keep API keys out of the browser, and send finished HTML.

Workarounds are piling up in custom code

Count the embeds. Each Code Embed is capped at 50,000 characters, and Webflow's support team can't help troubleshoot custom code. A dozen pasted scripts sit outside code review and tests, and the person who understands them is a developer anyway. You're paying for code without getting a codebase.

You want a way out

Check the exit before you need it. Code export is only on paid Workspace plans (Core is $19 a month billed yearly) and produces static HTML, CSS, JavaScript and assets. It leaves out CMS functionality, form processing, site search, code components and every locale except the primary one. CMS collections export separately as CSV, and exported code "can't be reimported". Leaving Webflow is a rebuild with the content carried over. That's a fair trade, as long as it's a known one.

What does moving from Webflow to Next.js cost?

More than the platform fees suggest. The bill has five parts.

The rebuild. Export won't give you a Next.js site, so templates, components and the content model are rebuilt. Webflow's DevLink exports native Webflow components to React, which keeps the design consistent but isn't a migration. For scale: we quote 3 to 5 weeks for a standard B2B marketing site; an app layer is scoped on top.

Design changes go through developers. Teams underestimate this one. A new section layout becomes a ticket, a pull request and a deploy. Editors still publish content themselves, and Sanity includes Visual Editing on all plans, but they edit inside components a developer built. We live with this: when we left Framer, design changes became code changes, while content edits stayed self-serve.

A separate CMS. Sanity Free is $0 with up to 20 seats, 10,000 documents and public datasets, and Sanity describes it as "for individuals and hobbyists"; Growth is $15 per seat a month with private datasets and 25,000 documents (Sanity pricing). Someone also has to design the content model. More in a CMS your team can run.

Hosting. Vercel says "Hobby teams are restricted to non-commercial personal use only", and a company's marketing site is commercial use. Vercel Pro is $20 a month with one deploying seat and $20 of usage credit. Elsewhere, Next.js runs as a Node.js server or in Docker with all features, or as a static export that drops redirects, rewrites, ISR and Server Actions.

SEO migration risk. Every old URL needs a destination. When we moved agr.studio off Framer, we ported about 80 redirect rules (old paths and renamed pages, English and Hebrew) and compared old and new sites page by page before switching the domain. From Webflow, work from CMS access, not just a crawl: it surfaces hidden pages and existing redirects. If you use Localize, rebuild hreflang on purpose: Webflow generates hreflang tags and routes locales with 302 redirects; in Next.js that becomes code you own. Keep content in server-rendered HTML, since Google notes "not all bots can run JavaScript". Our migration guide covers the full process.

A rough monthly platform bill

An estimate from list prices as of October 2026, excluding tax, build work and third-party tools. Assumptions: one English marketing site with a blog, one owner plus three content editors, under 50 GB of bandwidth.

• Webflow: Premium at $25 (billed yearly) or $39 (monthly), plus three limited seats at $15 or $19: about $70 to $96 a month, before any paid Workspace plan or Localize.
• Next.js + Sanity on Vercel: Vercel Pro at $20 plus Sanity Growth at $15 for each of four seats: about $80 a month, before usage beyond the included credit. On Sanity Free, it's Vercel's $20.

The fees land close together. The real difference is developer time: on Next.js, every design change and framework upgrade is paid in hours. No developer on staff or on retainer? Budget for one before you budget for the move, and read who owns your website after launch.

Can you keep Webflow for marketing and run Next.js for the app?

Yes, and it's often the right answer: marketing keeps shipping pages in Webflow, engineering owns the product in Next.js. The split makes sense when the two have different owners and release cycles, and the marketing pages don't need live data from the app. DevLink can bring Webflow components into the Next.js codebase so both share one design system.

Three ways to wire it:

• A subdomain. Marketing on company.com in Webflow, the app on app.company.com in Next.js. Cleanest ownership; the usual setup for logged-in products.
• A path, through Webflow Cloud. Webflow Cloud deploys Next.js and Astro apps from GitHub and can attach them to a Webflow site at a path (Webflow's example: my-site.webflow.io/my-app). It runs framework apps on Cloudflare Workers with partial Node.js support, and custom-domain mounting isn't available on the Starter or Basic site plans (overview). Test your dependencies against that runtime first.
• A path, through Next.js. Next.js rewrites can proxy to an external site, "especially useful for incrementally adopting Next.js" in the docs' words, including a fallback to your existing website after all Next.js routes are checked. Next.js owns the domain; everything else comes from Webflow. Good for moving section by section.

The split stops making sense when marketing pages need app data everywhere (per-account pricing, personalised content), or when you're paying for two platforms and two teams to run what is really one site. Then move the whole thing.

What have we seen on client projects?

Platform choice came up in 46% of the 101 client conversations we coded from January to October 2026, with Framer named in 34 of them. We've run migrations from Craft, Squarespace, WordPress, React and Webflow. One team with no Figma files ("the website is our design system") needed about two weeks per new page. A new platform wouldn't fix that; a CMS with documented components, planned to roughly halve the time, would.

Our own site is the other data point. In January 2026 we started rebuilding agr.studio on Next.js + Sanity, and it never launched. In October we rebuilt it on Astro + Sanity, because it's a content site: about 300 pages prerendered to static HTML, with only the embedded Studio and a few endpoints running on demand. Next.js can prerender too; Astro's static default simply fit a marketing site. If you're leaving Webflow for a pure marketing site, read Astro vs Next.js before you choose.

Which should you choose?

• Stay on Webflow if marketers ship most changes, there are no logged-in areas, and you fit within 20,000 CMS items and 10 locales.
• Split if the product needs code but marketing should keep its speed.
• Move to Next.js if the site is part of the product, the workarounds outnumber the native features, or engineering already runs Next.js and can own the site. See Next.js website development, and Framer vs Next.js for the same question from the Framer side.

If you're weighing the move, tell us what the site needs to do and we'll tell you which side of the line it's on.