Saleor Limitations: What Saleor Cannot Do in 2026

Saleor occupies a distinctive position in the headless commerce landscape: an open-source, BSD-3-licensed platform built on Python and Django, aimed at mid-market and enterprise teams, with a GraphQL-only API. Its own README states the philosophy plainly — "APIs are the only way to interact, configure, or extend the backend." For engineering-led organizations that want exactly that contract, Saleor is a credible, actively maintained choice, with 23.3k GitHub stars as of August 2026 and a steady release cadence.

That minimalism is also the source of Saleor's most predictable friction. Core covers catalog, order, and checkout primitives; almost everything else — payments, taxes, subscriptions, content, analytics, point of sale — is delegated to separately deployed apps or custom development. The official App Store listed 28 apps as of August 2026, and the cheapest commercial Saleor Cloud plan is $1,599 per month. Teams evaluating Saleor are therefore not really evaluating a finished product; they are evaluating a build program whose total cost depends on how many missing pieces their roadmap touches.

What follows is a factual inventory of what Saleor cannot do in 2026 — framed as architectural trade-offs rather than defects, with dates, prices, and sources wherever possible. Along the way, it notes how API-first platforms such as Swell approach the same problems with native functionality, often the most useful lens for deciding which side of the build-versus-buy line a given team should stand on.

Key Takeaways

• Saleor's API is GraphQL only — there are no REST endpoints for commerce data, so REST-expecting tools need custom middleware.
• Saleor Cloud has no entry-level commercial tier: managed hosting starts at $1,599 per month on the Select plan, with 0.8% GMV overage fees above $200,000 in monthly sales.
• Saleor has no native subscription or recurring-billing engine — a Saleor team response on GitHub states "Saleor doesn't support subscriptions" — and no subscription app exists in the official App Store.
• Multi-vendor marketplace functionality is absent; per a core maintainer, "Saleor was never designed as a marketplace platform," and Saleor's own marketplace solutions page describes building vendor management and split payments yourself via apps and Stripe Connect.
• The official App Store lists only 28 apps as of August 2026, leaving POS, ERP, loyalty, reviews, and subscriptions as custom builds.
• Self-hosting means operating at least six services — Django API, PostgreSQL, Redis/Valkey, Celery, SMTP, and object storage — and only versions 3.21 and newer receive security support.

A GraphQL-Only API and the Middleware It Demands

One Endpoint, One Query Language

Saleor exposes a single /graphql/ endpoint over HTTP POST, and that is the entirety of its queryable API surface. The official README calls the platform "GraphQL only," and the API documentation describes no REST alternative for reading or writing commerce data. Outbound webhooks do deliver JSON over HTTP, so event notifications can reach REST-shaped receivers — but inbound, every ERP connector, iPaaS recipe, or internal tool that expects REST endpoints needs custom middleware to translate. GraphQL is a genuinely good fit for storefront queries; the trade-off is that exclusivity narrows the range of off-the-shelf tooling that can talk to the platform at all.

API-first does not have to mean single-protocol. Swell's unified Backend API — the same API that powers Swell's own dashboard and checkout — provides full CRUD access to every data model with secret key authentication, and it is available through both REST and GraphQL access points, with a Frontend API offering partial CRUD access for browser contexts. The platform overview for developers and the developer tools documentation cover both surfaces.

Rate Limits, Query Costs, and Pagination Caps

Saleor Cloud sandboxes are limited to 120 API requests per minute per environment — applied to the whole environment, not per IP — returning HTTP 429 with a 60-second reset when exceeded. Production Cloud environments and on-premise deployments have no rate limiting, but two structural limits apply everywhere: queries exceeding the server's maximum GraphQL cost are refused, and paginated fields return at most 100 items per page. Every paid Cloud plan additionally carries a "fair API usage policy." None of this is unreasonable, but bulk operations — catalog imports, order exports, nightly syncs — must be engineered around cost ceilings and pagination loops from day one.

Hosted Pricing That Leaves Smaller Merchants Out

A $1,599 Monthly Floor for Commercial Use

Saleor Cloud's sandboxes are "Forever Free" — but restricted to non-commercial prototyping and testing. The moment a store takes real orders on managed hosting, the cheapest option is the Select plan at $1,599 per month, supporting up to $200,000 in monthly GMV. Volume is $3,999 per month up to $1,000,000 in monthly GMV, and Enterprise is custom-priced. For a smaller merchant, that is a roughly $19,000-per-year minimum for managed Saleor — before optional onboarding add-ons such as the $6,000 Accelerator or the $12,000 Forward Deployed Engineering package (both credited back as monthly credits in year one). The only alternative is self-hosting the open-source version, examined below.

Overage Fees That Rise With Revenue

The flat monthly fee is not the whole bill. Saleor Cloud charges GMV-based overage fees on orders above each plan's cap: 0.8% on Select, 0.4% on Volume, and "as low as 0.2%" on Enterprise. A Select merchant processing $300,000 in monthly GMV pays $1,599 plus roughly $800 in overage on the $100,000 above the cap — effective cost scales with revenue despite the flat-fee positioning.

Revenue-scaled pricing is common across commerce platforms, and Swell is no exception: its plans use revenue-based fee structures that apply only above each plan's revenue threshold, with per-order overage fees beyond that ceiling. The relevant difference for evaluators is what the fee buys — Swell is a managed platform whose feature set natively includes much of what Saleor treats as custom development, from subscriptions to B2B tooling.

The Operational Cost of Self-Hosting and Upgrading

Six Services Before the First Order

Self-hosting free Saleor means operating a nontrivial distributed system: the Django API itself, PostgreSQL, a Redis or Valkey cache ("Valkey is recommended," per the configuration docs), a Celery broker plus workers and exactly one Celery Beat scheduler, an SMTP gateway, and — typically — S3 or Google Cloud Storage for media. The Dashboard and storefront are separate applications deployed on their own; the docs FAQ confirms the platform is split across three separate repositories that ship together only via a docker-compose setup intended for local development. This is precisely the operational burden the $1,599-per-month Cloud fee exists to remove — self-hosting merchants carry it themselves, on-call rotations included.

Narrow Support Windows and Breaking Minor Upgrades

Saleor's security policy supports only versions 3.21 and newer; everything older receives no security patches. Minor upgrades routinely carry breaking changes documented in mandatory migration guides: the 3.21-to-3.22 guide requires first migrating through the latest 3.20.x release with a Celery worker enabled — "otherwise, you'll need downtime" — permanently removes five legacy Dashboard feature flags, deprecates GraphQL attribute fields, changes checkout-completion behavior, and relocates /apps/ paths to /extensions. Three minor branches (3.21, 3.22, and 3.23) receive concurrent patches, often several per week, with 3.23.29 released on August 26, 2026. Staying supported is an ongoing engineering commitment.

Managed platforms move that work off the merchant's plate entirely. On Swell, hosting, scaling, and security patching are the platform's responsibility, not the merchant's — teams needing implementation help can start with Swell's experts.

Recurring Revenue Without a Subscription Engine

No Native Subscriptions, by Design

Saleor core contains no subscription objects, billing cycles, or recurring order generation. The current payments documentation lists "recurring subscription logic" only as an example of what developers can build on the backend transactions API, and a member of the Saleor team put it directly in GitHub discussion #9359: "Saleor doesn't support subscriptions," recommending a custom app that pairs Stripe with Saleor metadata. As of August 2026, the official App Store's 28 apps include no subscription app of any kind. A merchant whose roadmap includes recurring revenue is signing up to design, build, and operate a billing system.

What a Custom Billing Stack Actually Involves

That build is larger than it first appears: recurring charge scheduling, failed-payment retries, secure card storage, plan changes, trial logic, and the interplay between recurring and one-time purchases all have to be solved — then maintained through every payment app update and platform upgrade.

This is one of the sharper contrasts with platforms that treat subscriptions as core commerce. Swell includes native subscriptions on all plans: physical or virtual products sold on flexible billing intervals, separate invoicing and fulfillment schedules, automatic payment retry (dunning), trials, and mixed carts combining subscription and one-time items in a single order — all managed from the same dashboard used for everything else. Subscription payments run through an encrypted card vault directly with the payment gateway, and merchants configure everything from subscription plans to ongoing subscription management without custom infrastructure.

Storefront and Checkout as Permanent Engineering Projects

An API-Only Checkout With No Hosted Option

Saleor provides checkout strictly as an API. There is no hosted, themeable checkout page, no theme store, and no drag-and-drop storefront builder; merchants build and host all customer-facing UI themselves. The official storefront repository offers a "minimal, production-ready storefront template" on React 19, Next.js 16, and TypeScript that works as a reference implementation and a starting point, with Stripe and a dummy gateway prewired — but branding, hosting, additional gateways, and every custom feature remain the merchant's to own indefinitely.

Swell approaches the same freedom from the opposite direction: its Checkout API lets teams build fully custom checkout and payment flows using the same APIs that power Swell's hosted checkout — making a ground-up build an option rather than a prerequisite for going live.

Payments Delegated to Apps — and the Stripe App's Gaps

Saleor core processes no payments itself; every gateway runs as a separately deployed app, and version 3.22 deprecated the remaining built-in payment and tax plugins (Braintree, Razorpay, SendGrid, the Avalara plugin, and the dummy gateways), with removal planned in future versions — pointing affected merchants toward app-based replacements such as the Juspay Hyperswitch, SMTP, and AvaTax apps. Even the flagship Stripe app has documented gaps as of August 2026: "Stripe Checkout is not currently supported," the method list stops at eight payment methods (ACH Direct Debit, Apple Pay, Card, Google Pay, Klarna, PayPal, SEPA Direct Debit, and Link — no iDEAL), one authorized amount can only be captured once, and existing app configurations cannot be modified.

Swell abstracts payments differently: a payment abstraction layer integrates with Stripe, PayPal, Braintree, Authorize.Net, and Amazon Pay, with PCI compliance handled by the platform — gateway breadth that matters especially for merchants in restricted or high-risk categories, as covered in the payments documentation.

Marketplace Ambitions Meet a Hard Architectural Ceiling

Not Designed for It — and Still Not Built In

On multi-vendor marketplaces, Saleor's maintainers have been unusually candid. In GitHub discussion #10641 (September 14, 2022), maintainer maarcingebala wrote: "Saleor was never designed as a marketplace platform, so it won't work out of the box" — and native multi-vendor features have not shipped since; Saleor's own marketplace solutions page describes building vendor management, commissions, and split payments yourself via attributes, apps, and Stripe Connect rather than any built-in capability. Nothing marketplace-related appears in the 3.23 release of April 2026 or in the current App Store. There are no vendor accounts, no vendor-facing admin, and no commission or split-payment handling anywhere in core.

The True Cost of the Workarounds

The workarounds discussed in that thread amount to custom development — as timuric noted, operators "would need to create your own admin interface for your vendors" — and the one community extension raised was later described as abandoned. Each piece is an engineering project: vendor onboarding, per-vendor catalogs, commission logic, and payouts all become custom code the merchant maintains indefinitely.

For businesses where the marketplace model is the business, this is the decisive evaluation point. Swell builds multi-vendor marketplace capabilities with split payment functionality into the core platform, so vendor operations become configuration rather than construction.

B2B Selling Without a B2B Suite

Missing Company Accounts, Customer Groups, and Quotes

Saleor core has no company or business accounts, no customer groups, no quote or RFQ workflow, and no net payment terms. Maintainers acknowledged the gaps in GitHub discussion #13220 (June 2023), pointing to open RFCs for customer collections and promotion rules; as of publication, promotion rules have since shipped (rule-based order promotions arrived with the new promotions engine in Saleor 3.19), but the documentation still contains no customer-groups feature page and no company-account or quoting functionality. Wholesale sales motions that depend on negotiated pricing and payment terms have nothing native to build on.

Approximating Wholesale With Channels and Metadata

The maintainer-recommended pattern is a dedicated B2B sales channel combined with metadata and custom apps, with B2B pricing approximated through per-channel price lists and rule-based promotions. It can work, but every layer — account hierarchies, negotiated terms, quoting — is bespoke.

Platforms with native wholesale tooling collapse that work considerably. Swell ships B2B and wholesale features natively: customer group-based pricing, custom payment terms, and quotes and draft orders via API, with custom fields available on every data model for the account structures a wholesale operation inevitably needs.

A 28-App Ecosystem and the Builds It Leaves Behind

What 28 Apps Cover — and What They Cannot

As of August 2026, the official Saleor App Store lists 28 apps: eight payment apps (including one partner integration), two tax apps (AvaTax and TaxJar), five CMS integrations, ten automation apps (Algolia, Klaviyo, Google Merchant Center, Twilio Segment, and others), one analytics app, one agentic-commerce integration, and one onboarding utility. Everything outside those categories — ERP, PIM, loyalty, reviews, shipping-rate shopping — is custom development against the API and webhooks. Notably, the store currently contains no point-of-sale app in any category, and the docs describe no POS module; retail POS on Saleor is an integration surface, not a product.

By comparison, Swell ships 40+ built-in service integrations, including Klaviyo, Omnisend, Mailchimp, ShipStation, and Algolia, alongside native tax integrations with Avalara AvaTax and TaxJar.

Analytics and Content Live Somewhere Else

The Saleor Dashboard historically shipped without built-in analytics or sales reporting. Saleor's first-party analytics app, Pulse, was announced on July 16, 2026 and now appears in the App Store; it is positioned for Saleor Cloud, with no stated availability for self-hosted stores as of publication — self-hosted merchants needing revenue, order, or fulfillment reporting must still pipe data into external BI tools. Content follows the same pattern: Saleor offers attribute-driven pages but no blog engine, theme system, or visual page builder, and the App Store's five CMS integrations (Strapi, Contentful, DatoCMS, Builder.io, and Payload) exist precisely because rich content is expected to live in another vendor's system.

Swell pairs its built-in dashboard with a native Contentful integration for teams that want a headless CMS, and handles international selling natively with content localization in 170 languages and pricing across 230 currencies — capabilities content-heavy global stores would otherwise assemble from parts.

Frequently Asked Questions

Does Saleor have a REST API, or is it GraphQL only?

GraphQL only. The official README describes Saleor as "GraphQL only" and headless and API only, and the API documentation covers a single /graphql/ endpoint with no REST alternative for reading or writing commerce data. Outbound webhooks deliver JSON over HTTP, so event payloads can reach REST-style receivers, but any REST-expecting tool that needs to query or mutate Saleor data requires custom middleware in between.

Can I sell subscriptions or recurring-billing products on Saleor?

Not natively. The current payments documentation lists "recurring subscription logic" only as something developers can build with the transactions API, and a member of the Saleor team stated plainly on GitHub that "Saleor doesn't support subscriptions," recommending a custom app pairing Stripe with Saleor metadata. As of August 2026, the official App Store contains no subscription app. Teams for whom recurring revenue is central should budget for a full custom billing build — or weigh platforms such as Swell, where native subscriptions are included on every plan.

What does Saleor Cloud actually cost, and is there a cheap starter tier?

There is no low-cost hosted tier. Free sandboxes are restricted to non-commercial use; the cheapest commercial plan is Select at $1,599 per month (up to $200,000 in monthly GMV, then 0.8% overage), Volume runs $3,999 per month (up to $1,000,000 in monthly GMV, 0.4% overage), and Enterprise is custom-priced with overage fees "as low as 0.2%."

The alternative is self-hosting the free open-source version, which means operating PostgreSQL, Redis or Valkey, Celery workers, SMTP, and separate Dashboard and storefront deployments in-house — a trade of subscription fees for infrastructure and staffing costs.

How much of the storefront and checkout do I have to build myself?

All of it, ultimately. Saleor's checkout is API-only with no hosted checkout page, and the open-source Next.js 16/React 19 storefront template is positioned by Saleor as a reference implementation and a starting point — cart, checkout, auth, and Stripe come prewired, but branding, hosting, additional payment gateways, and custom features are the merchant's responsibility.

Note also that the official Stripe app supports eight payment methods, including Apple Pay and Google Pay, but does not support Stripe's hosted Checkout and allows only a single capture per authorization — constraints worth validating against a store's payment requirements before committing.

Can Saleor power a multi-vendor marketplace?

Not out of the box. A core maintainer wrote that "Saleor was never designed as a marketplace platform, so it won't work out of the box," and native multi-vendor features have not shipped since — Saleor's own marketplace solutions page describes building vendor management, commissions, and split payments yourself via attributes, apps, and Stripe Connect. The workarounds discussed on GitHub amount to custom development, including a self-built vendor admin, with the one community extension raised in that thread later described as abandoned. As of August 2026, nothing marketplace-related has shipped in the 3.22 or 3.23 releases or appeared in the official App Store.