Squarespace to Code: How to Migrate to Astro, Next.js, or Payload CMS

The export is the whole argument

Most platform migration guides open with a case for leaving. This one does not need to, because Squarespace documents the case itself.

Squarespace offers one export path: Settings, then Advanced, then Import / Export. It produces a single WordPress-oriented XML file. Per Squarespace's own support documentation, that file contains layout pages, one blog page including its posts and up to 1,000 comments per post, text blocks, image blocks, the text from embed and Instagram blocks, and gallery pages. That list is exhaustive.

Everything else is named as excluded: album, cover, index, info, calendar, portfolio and store pages; page-specific headers, footers and sidebars; every blog page after the first; dropdowns; audio, video and product blocks; drafts; style settings; and custom CSS.

Read those two lists next to each other and the shape of the problem is clear. A portfolio site exports without its portfolio. An events business exports without its calendar. A store exports without its store pages. Every site exports without its design, because style settings and custom CSS are both on the exclusion list, and page-specific headers, footers and sidebars go with them. If you run more than one blog, one of them comes with you and the rest do not.

And then the fact that settles it. Squarespace states plainly that it is not possible to export content from one Squarespace site and import it into another.

That sentence is worth more than any argument we could make. The export file cannot round-trip inside the platform that produced it. It is not a backup, and it is not a portability guarantee. It is a partial WordPress importer, aimed at one specific destination, that drops most page types on the way.

What this means before you touch a destination

The practical consequence is that a Squarespace migration is a re-extraction project first and a rebuild project second. The XML is a convenience for your written text. The live rendered site is the actual source of truth.

That reverses the usual order of operations. Do not export and then plan. Crawl the site, list every page and page type, and check that list against what the export carries. The gap between the two is your real scope, and on a portfolio or commerce site the gap is most of the site.

It also puts a deadline on the work. Assets live on Squarespace's CDN, and those URLs stop being dependable once the subscription lapses. Download everything while the site is still up.

Products are a separate, also lossy, export

If you ran Squarespace Commerce, products do not travel in the XML. Store pages are on the exclusion list. Products have their own CSV export instead, capped at 10,000 per run, and Squarespace states that CSV omits additional product info, reviews, featured images, alt text and variant images.

So a store migration reconstructs product imagery, alt text, variant images and long-form descriptions from the live site, then rebuilds checkout on a dedicated service such as Stripe or a headless commerce platform. This is reliably the most expensive single piece of a Squarespace migration, and it is expensive for a documented reason rather than a vague one.

One cost note while you are still deciding. Squarespace charges a 2% commerce transaction fee on Basic, and 0% on Core, Plus and Advanced. If you are on Basic and selling, that fee is part of the comparison.

What Squarespace costs to stay on

As of July 2026, Squarespace plans start from $19 per month on annual billing, following a US price increase in the spring of 2026. Plenty of secondary sources still quote the older $16 entry price, and Squarespace also runs pricing tests, so check your own account and your own region rather than trusting a table in any article, including this one.

The subscription is rarely the deciding number anyway. It is a normal price for a hosted product. What the subscription buys you is a site whose export cannot be imported into another Squarespace site, which is the cost that only shows up on the day you want to leave.

The AI access gap is real here, specifically

We are cautious about the claim that closed platforms cannot be edited by AI, because for most of them it is now false. Webflow ships an official MCP server. Wix ships one. The WordPress core AI team ships an MCP adapter, with an Abilities API landing in Core 6.9.

Squarespace ships none as of July 2026. Third-party MCP servers built on Squarespace's public APIs do exist, so this is a gap in first-party support and coverage rather than a total blackout. Worth adding, since the naming invites it: developer.squareup.com belongs to Square, a different company. It is not a Squarespace developer platform.

The deeper difference is not the connector anyway, it is the artifact. Against a repo, an agent reads the entire site, changes it, runs the build to check its own work, and a bad change is one git revert away. Against a hosted builder API, it edits fields it can reach, with no build to verify against and no version history to roll back.

Choosing a destination

The destination matters less for price than people expect and more for fit.

Astro, for content sites

If your Squarespace site is a marketing site, blog or portfolio, Astro is usually the right answer. It is built for content-driven sites, ships zero JavaScript by default, and models each post as a Markdown or MDX file in a content collection. Images you host yourself get optimized at build time. For the majority of Squarespace sites this is the simplest landing spot.

Next.js, when the site is really an app

Choose Next.js when the site does more than present content: gated areas, dashboards, search, serious commerce, genuinely interactive features. Pages become React components, and a data layer backs the dynamic parts. It ships more JavaScript than Astro, which is the correct trade only when you need the interactivity.

Payload CMS, when editors need an admin UI

If non-technical people edit content daily, pair the frontend with Payload. Pages and collection items map to Payload Collections with typed fields, references and rich text, and editors get drafts, autosave and live preview. It is TypeScript end to end, and schema, content and frontend live in one repo.

Two things a technical buyer should know before choosing it. Payload was acquired by Figma in June 2025. And Payload Cloud has paused new deployments, so there is no first-party hosting product to sign up for. The software is MIT licensed and free. Hosting and a database are a separate bill, realistically $8 to $30 per month for a client-grade database.

What the running costs look like afterwards

Expect $0 to $20 per month on a static-first stack with Cloudflare Pages, or roughly $20 to $50 per month on Vercel with a managed database.

One warning, because it catches agencies and freelancers regularly. Vercel's Hobby tier is non-commercial only, and Vercel's own definition of commercial usage covers a paid employee or consultant writing the code, and receiving payment to create, update or host the site. A client site built on Hobby violates the terms. Vercel Pro at $20 per user per month is the floor for client work.

When you should stay on Squarespace

If your site is a handful of pages, you have no developer relationship, nobody is asking for anything the templates cannot do, and you are not selling at volume, stay. Squarespace is a well-built product and the export limits only bite on the day you leave. If that day is not visible from here, the migration is a cost with no return.

The trigger to move is usually one of four things. You need something the block system will not do. Performance is holding back a channel you can measure. You want the site in a repo so it can be changed by developers and agents rather than only through the editor. Or you have realised that the longer the site grows inside Squarespace, the more of it lives in page types the export refuses to carry.

The playbook

Crawl the live site and inventory every page type before exporting anything. Run the export, then diff it against the inventory so the omissions are on paper. Re-extract everything the XML drops, prioritising portfolio, calendar, index, album and store pages, since those are excluded wholesale. Export products to CSV if you sell, and reconstruct the imagery, alt text and variant images the CSV omits. Download every asset off the Squarespace CDN while the subscription is live. Rebuild content and design in your destination. Then map every old URL to its new path with 301 redirects and verify that map before cutting DNS.

The redirect map is the one step where a mistake is expensive and irreversible in the short term. Everything else can be fixed after launch.

Where to go next

If your site is a content or marketing site, read why Astro is the best framework for marketing websites, then the SEO migration survival guide. If non-technical editors must own content, the Payload playbook in our Webflow guide applies almost directly, since the destination work is the same regardless of what you are leaving.

If you are not sure how much of your site the export will actually carry, that is what a free migration review is for. We look at the real site, list the page types the XML will drop, and tell you honestly whether Astro, Next.js or Payload is the right landing spot.