By Ak · October 11, 2026
For someone who wants a portfolio, a landing page, a small interactive tool, or a page selling a few products, the traditional website setup can feel excessive. There is a builder to choose, hosting to configure, a database to connect, analytics to install, and another subscription whenever the requirements expand.
ChatGPT Sites changes that starting point.
My view is that it can make several of those purchases unnecessary for a meaningful group of users. Claims that it replaces every website platform are premature, but dismissing it as merely another website builder misses the change.
What Sites brings together
OpenAI describes Sites as a way to create and host websites, web apps, and games from a prompt or compatible existing project. It supports structured storage through D1, file storage through R2, public sign-in, custom domains where available, and automatic visitor and page-view analytics. The documentation also allows for external identity providers; describing it as limited exclusively to ChatGPT login is too restrictive. OpenAI Sites documentation
The benefit is practical: a user can start with an outcome instead of assembling services.
An event page does not necessarily need a website-builder subscription. A small tracker does not necessarily need a separately provisioned database. A portfolio does not necessarily need a separate hosting account. The infrastructure still exists. OpenAI manages more of it on the user’s behalf.
Which tools become less necessary?
| Existing purchase | My assessment |
|---|---|
| Hosting for a simple website | A strong replacement within Sites’ supported runtime and limits. |
| A basic landing-page or portfolio builder | A strong substitute when the generated result meets the owner’s needs. |
| Lovable or another AI app builder | Direct competition for overlapping projects; replacement depends on workflow and portability. |
| Notion Sites or a published Craft page | A plausible substitute for the public website, while the document-editing workflow may remain valuable. |
| A separately hosted database | Often unnecessary for a focused app whose requirements fit the included storage. |
| An analytics subscription | Potentially unnecessary for basic traffic counts; broader marketing analysis needs more. |
| A single-page course storefront | Plausible with an external commerce service; paid access and delivery still need a working process. |
| Shopify for a very simple shop | Plausible for some sellers; product count alone does not decide it. |
| Claude Code | Sites may remove a user’s reason to adopt a separate coding tool for one project, but it does not replace its wider development role. |
These are assessments of substitution, rather than claims that competing companies or categories disappear.
Lovable documents code downloads and two-way GitHub synchronization, including deployment outside Lovable. That is a concrete reason a builder might choose it even when Sites can produce the same visible app. Lovable GitHub documentation
Notion’s value includes its familiar page editor and database-backed content management. A person publishing documents they already maintain in Notion may prefer that workflow to prompting changes to a custom site. Notion Sites
The small-Shopify-store argument deserves more credit
Leaving Shopify does not necessarily mean building checkout, refunds, inventory, shipping, and taxes from scratch. For a sufficiently simple business, a different setup may be enough:
A Sites product page → an external hosted checkout → a simple fulfillment process.
Stripe Payment Links provides hosted payment pages without custom checkout code. Its documented capabilities include receipts, refunds, address collection, shipping charges, subscriptions, and tax integration. A seller using these capabilities would not be building every payment function themselves. Stripe Payment Links
That makes Sites credible for some small sellers: it can supply the storefront while another service handles payment.
However, “two to five products” is an incomplete measure of complexity. Two products with multiple sizes, limited stock, international shipping, and frequent returns can require more operational support than fifty downloadable files.
Shopify provides a coordinated workflow for orders, refunds, shipping configuration, fulfillment, and shipping labels. Those functions can justify its cost even for a small catalog. Availability of particular shipping services depends on eligibility. Shopify order management and fulfillment
My judgment is that Sites creates a credible alternative for shops with simple operations. It becomes a less obvious replacement as stock control, delivery, returns, and automation become central to the business.
A course sales page is easier than a course business
A persuasive page describing a course is one requirement. Taking payment is another. Reliably giving the correct buyer access—and withdrawing that access when appropriate—is another.
For a downloadable product or manually delivered workshop, a simple setup may be sufficient. For a course with protected lessons, student accounts, progress tracking, subscriptions, and access changes after refunds, more integration and testing are required.
Sites can reduce the website work. It does not automatically supply every course-management process.
A merchant-of-record service may handle part of the commercial burden. Lemon Squeezy explicitly documents that role, including responsibility for payments and sales-tax compliance within its service. The seller still needs to understand what the service covers and whether the product is eligible. Lemon Squeezy merchant-of-record documentation
Services should not all be grouped together. Stan’s documentation describes creator-configured tax collection through Stripe; that does not establish that Stan assumes the same merchant-of-record responsibilities as Lemon Squeezy. Stan tax documentation
Supabase can become unnecessary without becoming obsolete
If an app needs saved submissions, bookmarks, scores, or a small collection of records, included storage may remove the need to open a Supabase account. That is a real replacement of a purchase.
But database size alone does not determine suitability. Access controls, recovery, integrations, and developer workflows matter too.
Supabase documents Postgres, row-level security, multiple authentication options, realtime features, backups, and point-in-time recovery options. Those are capabilities to evaluate against an app’s requirements. Having paying users does not automatically mean an app must use Supabase or a separate backend. Supabase features
The costs do not all disappear
Sites usage is included up to plan-specific beta limits. Reaching those limits can restrict new Sites, additional storage, or continued public availability for a high-usage Site. Commerce owners remain responsible for their transactions and associated obligations. OpenAI Help Center
The economic comparison needs to include the additional ChatGPT cost attributable to the website, domain renewal, payment-provider charges, other services, maintenance time, and the cost of moving elsewhere if requirements change.
For someone already paying for ChatGPT, the additional subscription cost may be small or zero within the included limits. That makes Sites especially competitive. Removing a website-builder bill does not remove payment processing costs or the work of operating a business.
One unresolved issue: the payment documentation conflicts
The current OpenAI Help Center explicitly permits selling goods or services through a third-party payment processor. The Sites developer guide still lists enabling financial transactions among unsupported uses.
Both statements remain visible. The Help Center supports the external-checkout approach, but the documentation is inconsistent; it would be inaccurate to describe commerce support as completely settled. OpenAI Help Center, Sites developer guide
My view
The strongest part of this change is how much less a person needs to decide before starting.
Previously, “I need a small website” could become a shopping exercise across builders, hosts, databases, and integrations. Sites can turn more of that into a single conversation.
My inference is that this puts substantial pressure on tools bought mainly to publish a simple page or host a focused app. Specialized platforms retain stronger reasons to exist when their value comes from editorial workflows, commerce operations, recovery, integrations, or control over the code.
Some sellers genuinely do not need everything Shopify offers, and an AI-built storefront with hosted checkout may suit them.
ChatGPT Sites can make much of the traditional website setup optional for simple projects. It gives small businesses more ways to choose—and pay for—only the services they actually need.