
-
Shopify Plus
+2
Shopify B2B
Shopify B2B in 2026: The Complete Guide to Wholesale on Shopify
The honest 2026 Hydrogen assessment
@shopify/hydrogen@2026.4.0, released April 9, 2026. Calendar versioning (year.release.patch) replaced SemVer in 2024.The Shopify Hydrogen landscape in 2026 looks very different from the one most articles on the internet describe. Hydrogen ships on a calendar cadence now. Oxygen is still the default runtime and still free with your Shopify subscription. Storefront MCP went live in Winter '26 and opens up a new class of AI-native storefront capabilities. Allbirds, SKIMS, Gymshark, and Good American are all running on it at scale.
At Ask Phill we have built on Liquid, Hydrogen, Next.js + Storefront API, and Remix-over-Storefront-API. We do not push Hydrogen to every client who asks about headless. We also don't dismiss it. This guide is the honest version of the conversation we have with brands considering Hydrogen today: what it actually is in 2026, when it is worth the investment, when it isn't, and what it will cost you over three years.
If you are still deciding whether to go headless at all, start with our headless Shopify decision guide. This article assumes you are serious about headless and trying to decide which framework and runtime to bet on.
Two products, two different things, frequently confused. Before going further it's worth being precise.
Hydrogen is Shopify's open-source React framework for building custom storefronts against the Storefront API. It is distributed as the @shopify/hydrogen npm package. It is React Router based (since Shopify's 2022 acquisition of the Remix team, Remix itself has merged into React Router 7). It is free. It runs on any Node or Cloudflare-compatible runtime, not only Oxygen.
Oxygen is Shopify's edge hosting runtime built on Cloudflare's open-source workerd project. It's where most Hydrogen storefronts are deployed, because it is included at no additional cost with every Shopify plan, provides unlimited bandwidth, and auto-scales. Oxygen is not a requirement for Hydrogen. You can deploy Hydrogen to Vercel, Cloudflare Workers directly, Netlify, or self-host. But Oxygen is the only runtime with deep, first-party Shopify integration.
Both Hydrogen and any other headless frontend (Next.js, Nuxt, custom) read data from Shopify through the Storefront API. Hydrogen is the React framework that wraps this API with opinionated primitives (cart logic, customer accounts, analytics, checkout redirects, internationalization). It saves you from rebuilding solved problems.
Think of it this way: Hydrogen is the framework, Oxygen is the runtime, Storefront API is the data layer. You can swap any two of the three.
Hydrogen has been around since 2022. For most of its life it was interesting but early. That changed in the last nine months. If you last evaluated Hydrogen in 2023 or early 2024, your assessment is probably wrong now.
Shopify moved Hydrogen to calendar-based versioning. @shopify/hydrogen@2026.4.0 shipped on April 9, 2026. The current cadence is one headline release per quarter with patch releases in between. This is how stable infrastructure projects version. It is not how experimental frameworks version.
The September 2025 release landed React Router 7.9.2 integration with first-class type-safe routes and middleware support. Hydrogen now gives you the same routing primitives as any modern React framework, with the Shopify-specific commerce features layered on top.
The December 10, 2025 Winter '26 Edition added Storefront MCP support to Hydrogen storefronts on Oxygen. Model Context Protocol lets you expose your store's live product, cart, customer, and order data to AI agents built with any MCP-compatible tooling. The concrete implication: a custom AI shopping assistant running inside your storefront can read your actual catalog and cart state in real time rather than relying on a stale training snapshot.
This is not a feature flag or a demo. It is infrastructure-level integration. It is also currently an Oxygen-only capability, which changes the Oxygen-vs-alternatives calculus for teams that care about AI commerce.
The April 2026 release (2026.4.0) forced some meaningful breaking changes: the Storefront API proxy is always enabled (the proxyStandardRoutes config was removed), backend consent mode is supported, Storefront API 2026-04 is the new baseline. Breaking changes in a mature framework are not a red flag. They are the signal that Shopify is willing to push the framework forward rather than maintain backwards compatibility at the expense of progress.
Allbirds is running Hydrogen across 35 countries and integrating with 50 retail locations. SKIMS, Gymshark, Good American, and Denim Tears are all live on Hydrogen. This is enough market validation to take the framework seriously.
From our migration work, Hydrogen is the right choice when these apply:
Hydrogen optimizes hard for Shopify-specific problems: cart edge cases, customer account flows, checkout redirects, Shopify analytics, B2B primitives, Shop Pay integration, Shopify Markets. If your three-year plan keeps Shopify as the core commerce system, Hydrogen's opinionated integration saves you from rebuilding that surface area in your own code.
Oxygen is free, global, unlimited bandwidth, and integrated with GitHub for CI/CD out of the box. If you do not have a dedicated DevOps function, this is a significant operational win. The alternative is running Vercel or Cloudflare Workers yourself, which means paying for bandwidth and managing the deployment pipeline.
Hydrogen is a React framework. Your team (or your agency) needs senior React developers who understand server-side rendering, data loaders, and the React Router model. If your in-house team is Liquid-only or PHP-only, the hiring cost is real.
Hydrogen shines for one brand, one storefront, high traffic, meaningful custom UX. It can do multi-brand, but the ergonomics are best when you have one primary domain and a single coherent customer experience.
Storefront MCP is Oxygen-only today. If your roadmap includes custom AI shopping assistants, personalization agents, or conversational commerce, Hydrogen on Oxygen is currently the best-integrated path on the Shopify side. You can build similar experiences on Next.js + Storefront API, but you'll rebuild the MCP bridge yourself.
We talk a meaningful percentage of brands out of Hydrogen when they ask. Common reasons:
If your architecture treats commerce as one of many services (commercetools for catalog, Algolia for search, Contentful for content, Shopify for checkout only), Next.js + Storefront API is usually a better fit. Hydrogen assumes Shopify is the center. Next.js assumes nothing.
Hydrogen does not solve the team-size problem. A small team shipping fast on Liquid is almost always going to beat that same team trying to ship a first Hydrogen build. Unless you already have Hydrogen experience in-house, the learning curve will dominate your first-year velocity.
If you operate multiple distinct brands with independent teams, independent catalogs, and independent fulfillment, a single Hydrogen build with Shopify Markets may not be the right tool. You may need multiple storefronts, or a different architectural pattern entirely. Talk to a Shopify Plus Partner before assuming Hydrogen is the answer.
Oxygen is opinionated. It runs on Cloudflare workerd. It has its own logging, its own observability, its own deployment pipeline. If your engineering org has standardized on Vercel, AWS, or your own Kubernetes, the forced adoption of Oxygen's way can create friction. You can deploy Hydrogen to other runtimes, but you lose Oxygen's zero-cost convenience and some first-party integrations.
Liquid is the right answer for most Shopify brands. Liquid themes today are fast, well-optimized, and give you access to everything Shopify builds into the platform: Theme Blocks, Hydrogen React primitives (yes, they work in Liquid too), Shop Pay, and the full native checkout. If your current storefront works, your brand doesn't need radical custom UX, and your conversion rate is healthy, stay on Liquid.
Go headless only when Liquid's rendering model genuinely blocks something that matters to your business. Common legitimate reasons: complex interactive product configurators, real-time personalization that Liquid can't express efficiently, heavy app-like UX with deep client-side state, or content composition that Liquid's templating constrains.
Next.js has the larger ecosystem, more flexible deployment options, and a larger talent pool. Choose Next.js when you need maximum flexibility across a composable stack, heavy content routing, or deep integration with non-Shopify systems. The trade-off is rebuilding the Shopify-specific conveniences that Hydrogen gives you for free: cart handling, customer accounts, checkout redirects, Shopify analytics, B2B primitives, MCP integration.
Our rule of thumb: if Shopify is 80% or more of your commerce logic, Hydrogen. If Shopify is less than 50%, Next.js. In between, it depends on team and timeline.
Remix effectively merged into React Router 7, which is the foundation Hydrogen now builds on. A standalone Remix-over-Storefront-API build gives you more flexibility at the cost of rebuilding Shopify integration. In 2026, choosing Remix standalone over Hydrogen is mostly a preference for team autonomy over Shopify's opinions. If your team already has strong Remix experience and doesn't need Oxygen's free hosting, it's a reasonable choice.
If your team is Vue-first, Nuxt + Storefront API is the pragmatic option. You lose the MCP integration, the free Oxygen hosting, and the Shopify-specific primitives. You gain a framework your team already understands. For Vue teams, the team fit usually beats the framework fit.
| Criteria | Liquid | Hydrogen | Next.js + Storefront API | Remix standalone | Nuxt + Storefront API |
|---|---|---|---|---|---|
| Shopify-first integration | Native | Deep | Manual | Manual | Manual |
| Hosting cost | Included | Oxygen free | Vercel paid | Varies | Varies |
| Storefront MCP support | No (native AI via admin only) | Yes (Oxygen only) | Manual bridge | Manual bridge | Manual bridge |
| Team skill required | Liquid | Senior React | Senior React | Senior React | Senior Vue |
| Composable-stack flexibility | Low | Medium | High | High | High |
| Editorial/content flexibility | Medium | Medium | High | High | High |
| Time to first build | Fast | Medium | Slow | Slow | Slow |
| Three-year TCO (typical) | Lowest | Medium | Medium to high | Medium to high | Medium to high |
You have four realistic options. Each has trade-offs.
Runtime: Cloudflare workerd, edge global Cost: €0, included with all Shopify plans Bandwidth: Unlimited CI/CD: GitHub integration out of the box, environment management, preview deployments Storefront MCP support: Yes, Oxygen-only Best for: Teams who want zero-config deployment and minimum infrastructure overhead. 90% of Hydrogen builds we deliver run on Oxygen.
Runtime: Node + Edge Functions Cost: Pay per request, typically €100 to €500+ per month depending on traffic Bandwidth: Included up to tier limits, then pay-as-you-go CI/CD: Vercel's pipeline, excellent preview environments Storefront MCP support: Not native, requires manual bridge Best for: Teams already deep in the Vercel ecosystem, brands running Next.js for non-commerce pages and wanting deployment consistency.
Runtime: workerd (same as Oxygen but self-managed) Cost: Workers Paid plan at $5/month + request-based billing. Bandwidth free within Cloudflare network. CI/CD: Wrangler CLI + whatever you build around it Storefront MCP support: Not native, requires manual bridge Best for: Engineering orgs that already run on Cloudflare and want one unified infrastructure story.
Runtime: Node.js on your infrastructure Cost: Whatever your infrastructure costs Bandwidth: Your responsibility CI/CD: Your responsibility Storefront MCP support: Not native, requires manual bridge Best for: Large engineering orgs with existing platform investments and regulatory constraints that prevent edge hosting.
For almost every brand we work with, Oxygen is the right call purely on the total cost of ownership math. Vercel makes sense if you have a strong existing reason to stay there. The other two are niche.
The framework and hosting are free. The rest is where it gets real. All figures below are based on Ask Phill's client portfolio and are EUR, not Shopify-published numbers.
See our Shopify Plus pricing guide for the full TCO breakdown.
A Hydrogen storefront in production typically needs at least:
Whether that's in-house, via retainer, or a mix is the real decision. Plan for around €15,000 to €40,000/month in effective team capacity depending on level and location. For more on team structures, see our Shopify Plus Partner guide.
This is the most under-covered story in the Hydrogen space right now. It matters because it materially changes the competitive position of Hydrogen-on-Oxygen versus any alternative stack.
Model Context Protocol is an open standard that lets AI agents request live data from services via a structured interface. Storefront MCP is Shopify's implementation for commerce data. A Hydrogen storefront on Oxygen now exposes MCP endpoints that AI agents can read from. The agent sees your actual product catalog, current cart state, customer context, and order history in real time rather than relying on training data.
You can build all of the above on Next.js or Remix. You will rebuild the MCP bridge yourself. On Hydrogen-on-Oxygen, it's infrastructure-level integration. For AI commerce roadmaps, this is a concrete reason to choose Hydrogen over alternatives if you are Shopify-first.
The scope of what MCP will unlock is still emerging. Conversational commerce, agentic checkout flows, and AI-driven merchandising are all in early stages. If you believe this is where commerce is heading (we do), being on the platform Shopify is actively integrating with is strategically meaningful.
Most Hydrogen content skips the operational story. It is where most first-time Hydrogen projects underestimate effort.
React and Remix talent is expensive and in demand. In the Netherlands or the UK, a senior Hydrogen-capable developer is €75,000 to €110,000 loaded. You'll need at least one, realistically two. In practice, most mid-market brands use an agency retainer for the initial build and for the first 12 to 18 months of production, then consider in-housing once the codebase stabilizes and the product roadmap is clear.
Oxygen's GitHub integration is genuinely good. You connect a repository, you get automatic preview deployments per pull request, environment variables flow through, and production deploys are one click. If you deploy to Vercel or Cloudflare directly, you'll build this layer yourself. Plan accordingly.
Hydrogen on Oxygen gives you basic deploy logs and error reporting. For serious production monitoring (error rates, performance metrics by route, alerting, SLO tracking) you'll integrate Sentry, DataDog, Logtail, or similar. Budget the monthly cost and the setup time.
For any serious content workflow, your merchandisers need to preview changes before they go live. Oxygen's preview deployments cover code changes. For content changes via a CMS, you'll need to wire the CMS's preview mode to your Hydrogen dev server. Plan for one to two days of integration work per CMS.
This is the quietest cost of headless. Your merchandisers stop editing themes in the Shopify admin and start editing in the CMS (if you use one) or in code (if you don't). Navigation changes, homepage updates, promotional banners, and product page variations all move to a new workflow. If your team is two people, this is fine. If your team is 15, this is a project in its own right.
Public Hydrogen deployments across categories. This is not exhaustive. These are brands we can independently verify as running on Hydrogen.
Allbirds runs Hydrogen across 35 country storefronts integrated with 50 physical retail locations. The scale of the multi-region, omni-channel architecture is the clearest public validation that Hydrogen can handle enterprise complexity.
SKIMS was an early Hydrogen adopter and has scaled extensively on the platform. It remains a high-profile reference for Hydrogen's ability to handle drop-driven, high-traffic commerce.
Each represents a different use case: Gymshark for athletic apparel and global distribution, Good American for fashion and inclusive sizing workflows, Denim Tears for premium/streetwear with distinctive custom UX.
What these brands have in common: they are Shopify-committed, they have engineering capacity (in-house or via agency), and they needed UX flexibility that Liquid couldn't easily deliver.
We don't yet see many mid-market brands (€5-15M ARR) on Hydrogen. The cost ceiling and team requirements still skew enterprise. This will change as the framework matures and the talent pool grows, but today, Hydrogen is a choice that fits brands with either significant scale or strong engineering conviction.
At Ask Phill we recommend Hydrogen when:
We push back on Hydrogen when:
In our deal portfolio, roughly one in ten brands that arrive asking about Hydrogen actually leave with a Hydrogen build. The rest stay on Liquid, move to a Liquid-plus-app-extensions approach, or go composable on Next.js when the situation calls for it. This is the honest distribution, and it's why we don't sell Hydrogen as a default answer.
Yes. Oxygen is actively developed, included at no extra cost with all Shopify plans, and gained Storefront MCP support in the Winter '26 Edition release on December 10, 2025. Oxygen is the primary deployment target for Hydrogen storefronts and is the only runtime with native MCP support today.
@shopify/hydrogen@2026.4.0, released on April 9, 2026. Shopify moved Hydrogen to calendar-based versioning (year.release.patch), so the next scheduled release will follow a quarterly cadence.
No. Shopify moved Hydrogen to calendar versioning, so the version numbers now look like 2026.4.0 rather than 3.0. Calendar versioning is common for mature infrastructure projects and signals stable, predictable release cadence.
The Hydrogen framework is free and open source. Oxygen hosting is included at no extra cost with all Shopify plans. What costs money is the build (development time), the ongoing team to maintain it, and any third-party services (CMS, search, monitoring) you layer on top.
From Ask Phill's portfolio, a focused Hydrogen build on an existing Plus store runs €60,000 to €180,000. A full Liquid-to-Hydrogen migration with custom integrations runs €150,000 to €400,000. An enterprise composable build (Hydrogen + CMS + search + personalization) runs €300,000 to €700,000+. Ongoing team cost is €8,000 to €25,000/month depending on scope.
Yes. Hydrogen runs on any Node or Cloudflare-compatible runtime, including Vercel, Cloudflare Workers directly, Netlify, or self-hosted Node. The trade-off versus Oxygen: you pay for bandwidth and compute, you manage the deployment pipeline yourself, and you lose native Storefront MCP support (you can still build MCP bridges manually).
Storefront MCP is Shopify's implementation of the Model Context Protocol, released with the Winter '26 Edition in December 2025. It lets AI agents read live product catalog, cart, customer, and order data from Hydrogen storefronts running on Oxygen. This is the infrastructure layer for AI shopping assistants, personalization agents, and conversational commerce experiences that use real-time store data rather than stale training snapshots.
Not necessarily. A well-optimized Liquid theme with Shop Pay, CDN-cached resources, and modern assets often matches or beats a poorly built Hydrogen storefront. Hydrogen gives you more control over the performance ceiling, but it also gives you more ways to build a slow site if you're not disciplined. Performance is an outcome of engineering care, not of framework choice.
Stay on Liquid if your current storefront is functionally adequate and your drive for headless is aesthetic rather than measurable. Choose Next.js + Storefront API if your architecture is composable with Shopify as one of several systems. Stay on Liquid if your team is too small to sustain a React codebase or if your timeline requires first-build-fast and you have no Hydrogen experience in-house.
Yes, but it takes planning. The typical pattern is to start with a single Hydrogen route on a subdomain or path (for example, a product configurator or a specific collection page) running alongside your Liquid theme. Over time, additional routes migrate to Hydrogen until the full site is headless. The migration is more complex than a clean-slate rebuild because you're running two rendering stacks simultaneously, but it carries less launch risk.
As of 2026, well-known Hydrogen-powered brands include Allbirds (35 country storefronts and 50 retail locations), SKIMS, Gymshark, Good American, and Denim Tears. The common thread is Shopify-committed brands with engineering capacity and UX ambitions that go beyond what Liquid delivers most naturally.
A focused Hydrogen build on an existing Plus store is typically 2-4 months. A full Liquid-to-Hydrogen migration with custom integrations is typically 4-7 months. An enterprise composable build is typically 6-10 months. Add 4-8 weeks of paid discovery before the build itself for projects above €100,000.
If you are weighing Hydrogen for a Shopify Plus build, the right starting point is a paid discovery that maps your UX ambitions against your team capacity and three-year cost picture. Our Shopify development team handles Hydrogen discovery, build, and ongoing support. To scope a discovery conversation, get in touch.
Do you want to
stay updated?
Yes

Shopify B2B
Shopify B2B in 2026: The Complete Guide to Wholesale on Shopify

When to Choose Headless vs Native Shopify Plus
Shopify Headless Commerce: When It's Worth It, When It's Not, and What It Actually Costs

What the Shopify AI Toolkit actually is
How to Use the Shopify AI Toolkit: Setup, Tools & What You Can Build
Stay ahead
Subscribe to our newsletter for a roundup of the latest in ecommerce, straight to your inbox.