Shopify Hydrogen vs Liquid: What Each One Will Not Do

Shopify Hydrogen went generally available in the Fall '26 Edition on 28 October 2026, rebuilt to work with any framework and any runtime. That settles the old argument about which stack to pick, and it does not settle the decision in front of you.

The short answer: Liquid is the storefront Shopify edits for you, through the theme editor, Canvas and app embeds. Hydrogen is the storefront you build and maintain yourself, and Shopify says plainly that it costs more to build and run. Choose based on who maintains it, not on which is faster.

What Hydrogen actually is now

Hydrogen is Shopify's headless toolkit, currently on version 2026-10. It is not a framework. Your framework owns routing, rendering, data loading and UI. Hydrogen owns the Shopify-specific logic that is easy to get wrong: the Storefront API client, the cart, variant selection, collection filtering, analytics, and the routes that apps and checkout depend on.

According to Shopify's Hydrogen documentation, it splits into a framework-agnostic core plus bindings. React and Vue get packaged bindings. Astro, SvelteKit and SolidStart use the core directly. Supported frameworks include Next.js, React Router, Nuxt, Astro and SvelteKit.

On the server, Hydrogen's request handlers run around your router and answer the handful of routes a Shopify storefront needs, like /api/cart and /checkout, then pass everything else through. If your router cannot find a page, Hydrogen checks the store's URL redirects before a customer sees a 404.

What shipped in this edition: customer accounts, predictive search, Vue bindings, edge caching and agent skills, all generally available, plus Vercel Shop, an open-source Next.js template built on the Hydrogen SDK.

The three-layer split, in one table

That third row is the part people miss. Headless does not mean headless checkout. Checkout stays on Shopify in both architectures, which is why the buying experience you actually optimise rarely changes between the two.

What Liquid does that Hydrogen cannot

This is the honest asymmetry, and it got wider in October rather than narrower.

The Fall '26 Edition also shipped a new theme architecture: page structure now lives in the Liquid template itself rather than a separate JSON file, with blocks, typed contracts, partial updates, markdown files that guide agents, and Tailwind built in. Themes built this way are designed to be edited by Sidekick in Canvas.

None of that applies to a Hydrogen storefront. If you build headless you do not get the theme editor, app embeds, market-specific theme customisation, or theme updates pushed by a vendor. You also do not get Canvas, which as we covered in the Canvas comparison only works with Shopify and custom Liquid themes anyway.

One thing that is no longer a differentiator: agent readiness. WebMCP tools now sit on every Liquid and Hydrogen storefront and checkout, so browser agents can read structured data, update the cart and complete checkout in both. If someone tells you headless is required for AI agents, that stopped being true in October.

The four real limits of going headless

Shopify is unusually direct about this in its own headless overview: a headless storefront costs more to build and run than a theme, and the merchant needs developers to maintain it after launch. Read that as the vendor arguing against its own upsell.

1. Ongoing maintenance is yours. Framework upgrades, dependency patches and Hydrogen version migrations. Version 2026-04 and earlier now sit on a legacy track with a documented migration to 2026-10.
2. You lose the editing surface. No theme editor, no Canvas, no Sidekick on the storefront. Every copy or layout change becomes a deploy.
3. Two APIs and a secrets model to get right. Storefront API and Customer Account API, each with public and private tokens through the Headless or Hydrogen channel. Shopify warns explicitly never to expose the private token client-side or give it a browser prefix like NEXT_PUBLIC_. Tokenless access covers products, collections, search and cart, capped at 1,000 query complexity.
4. A higher floor for launch. Hydrogen's value is in cart, variant and search logic. If your storefront barely sells, you paid for plumbing you could have had for free.

The four limits of staying on Liquid

Symmetry matters, so here is the other side.

1. You are constrained to what Liquid and the platform expose. Genuinely novel interactions fight the platform.
2. Performance is a theme and app problem. The script weight on a Liquid storefront comes from themes and app embeds, and you inherit it.
3. The editing surface is the ceiling and the floor. Fast to change, but hard to change structurally without a developer.
4. Single surface. Themes do not extend to a native app, a kiosk or a game.

How to actually decide

We sell themes and an AI theme builder, so our interest is on the Liquid side and every line above should be read knowing that.

The decision rule that survives our bias is this: Hydrogen is a staffing decision, not a performance decision. Pick it when you already employ people who own a JavaScript frontend, you need channels Liquid cannot reach, or you must integrate a CMS, ERP or PIM into the storefront. Stay on Liquid when the people who maintain the store are merchants or marketers, and when the storefront is a web store rather than a product in its own right.

There is a middle path worth naming because we measured it in our own headless review for 2026: take the design discipline of headless without the architecture. Most stores that say they need headless actually need a faster theme.

7-day free trial · all themes included · cancel anytime