← Blog
Foundations

Build a ChatGPT Plugin (Nov 2026): MCP, Events, Review Steps

Thien Nguyen · Oct 6, 2026

To build a ChatGPT plugin in 2026, put a remote MCP server in front of your product, add OAuth 2.1 for user data, add optional MCP Apps UI, then upload a ZIP to OpenAI's plugin portal for review. A read-only plugin takes two to three engineer-weeks; one with UI and MCP Events, two to three months.

Updated 13 November 2026. "Plugin" is what OpenAI now calls a package of skills, an MCP server and optional UI that runs in ChatGPT and Codex. Since DevDay on 29 September 2026, plugins can also live in the ChatGPT sidebar, open panels beside a conversation, act as file viewers and start automations through MCP Events, and OpenAI's always-on dots reach your product through them. This guide covers the naming, the architecture, every build step, submission and review, cost and timeline, and what we found when we checked our own MCP servers against OpenAI's requirements. Every Apps SDK, directory and submission fact was re-checked on OpenAI's developer docs on 13 November 2026.

Is a ChatGPT plugin in 2026 the same as the old ChatGPT plugins?

No. The original ChatGPT plugins from 2023 were a beta built on an OpenAPI manifest, and OpenAI shut them down in 2024. Today's ChatGPT plugins are a different product: a package of skills, a remote MCP server and optional UI, published to the ChatGPT directory and also usable in Codex.

The naming changed three times, which is why search results are confusing:

PeriodWhat OpenAI called itHow it workedStatus in November 2026
2023 to early 2024ChatGPT plugins (beta)An ai-plugin.json manifest plus an OpenAPI spec of your REST APIShut down. New plugin chats ended on 19 March 2024 and existing ones on 9 April 2024, per OpenAI's wind-down notice
Late 2023 onwardGPTs with ActionsA custom GPT that calls your API through an OpenAPI schemaA separate product, not covered here
2025 to 2026Apps in ChatGPT, built with the Apps SDKA remote MCP server with optional UI componentsThe Apps SDK documentation now redirects to the plugins docs
2026Plugins for ChatGPT and CodexSkills, MCP servers and optional UI in one package, with extensions and MCP EventsCurrent. Documented at developers.openai.com/plugins

So if you search "how to create a plugin for ChatGPT", ignore tutorials that start with ai-plugin.json. Those describe the retired 2023 system. Everything in this guide follows the current plugin documentation.

What changed for ChatGPT plugins at DevDay 2026?

Plugins went from tools the model calls in the background to app-like features inside ChatGPT. At DevDay on 29 September 2026, OpenAI announced:

  • Plugin extensions. A plugin can have an entry in the ChatGPT sidebar, open interactive panels next to a conversation, and act as a viewer for your file types. OpenAI showed Figma and Adobe integrations.
  • MCP Events. ChatGPT supports the proposed MCP Events specification, so a plugin can start an automation when something happens in a connected app, such as a new task on a project board.
  • Updated creation, discovery and review. Plugin recommendations during conversations, review tracking with clearer feedback, and simpler updates to published plugins.
  • Dots. Persistent agents with connected apps and their own cloud computer. A dot uses the plugins a user has installed and enabled, with their existing permissions.

OpenAI's DevDay 2026 recap puts ChatGPT at 1.2 billion weekly users, its own figure. A listing is not the same as being found, as the review section explains, but the plugin is now how your product shows up when a customer, or their dot, works inside ChatGPT. For how dots compare with building your own agent, see OpenAI Dots, Meta Muse or a custom AI agent.

What is the architecture of a ChatGPT plugin?

A ChatGPT plugin is a package that contains skills, a remote MCP server, or both, with optional UI that the server returns. The MCP server is the core: ChatGPT, Codex and dots call its tools over Streamable HTTP, and it calls your API with the user's OAuth token. OpenAI's documentation describes four shapes:

ShapeWhat it containsChoose it when
Skills onlyFolders with a SKILL.md and supporting files that tell the model how to run a workflowInstructions plus tools ChatGPT already has are enough. No access to your product is needed.
MCP server onlyA remote MCP server exposing your product's tools, with authThe model needs live data or actions from your product.
Skills and MCP serverWorkflow instructions that use your MCP toolsUsers run multi-step workflows, such as "prepare the monthly report", on top of your tools.
MCP server with UIThe above plus UI resources rendered in ChatGPTPeople need to inspect, compare, edit or confirm structured data visually.

In words, the request path looks like this: the user (or a dot) asks ChatGPT something, the model picks one of your tools, ChatGPT calls your MCP endpoint with a bearer token, your server checks authorisation and calls your API, and the tool returns structured content. If the tool is linked to a UI resource, ChatGPT renders it in a sandboxed iframe that talks back over the MCP Apps bridge. For MCP Events, the direction reverses: your product POSTs signed webhooks to a callback URL ChatGPT gave you.

For a SaaS product, the MCP server defines which tools exist, their input and output schemas, who may call them, and what data they return. UI is optional, and OpenAI's guidance is to build the tools first and keep every tool useful without UI, so dots can run the same workflow headlessly. If you have not built an MCP server before, start with our MCP server development guide. It covers the protocol, OAuth 2.1 architecture and security in depth, so this article focuses on what is specific to ChatGPT.

ChatGPT plugin architecture: ChatGPT, Codex and Dots call a plugin package whose core is a remote MCP server with skills, MCP Apps UI, extensions and MCP Events, connected to your OAuth authorization server, your API and product events through signed webhooks
Figure 1: ChatGPT plugin architecture: ChatGPT, Codex and Dots call a plugin package whose core is a remote MCP server with skills, MCP Apps UI, extensions and MCP Events, connected to your OAuth authorization server, your API and product events through signed webhooks

How do you create a plugin for ChatGPT in 2026?

Build a remote MCP server with a few well-annotated tools, add OAuth and optional UI, test it in ChatGPT, then package it as a ZIP and submit it in the plugin portal. Nine steps take you from idea to a published listing:

  1. Pick three to five use cases. Write them as things a user would say in ChatGPT, such as "show me overdue invoices for Acme". These become your tools and, later, your review test cases.
  2. Build the remote MCP server. Expose a Streamable HTTP endpoint, typically at /mcp, on a stable public HTTPS domain. Keep a small set of tools with strict schemas and plain verb names like get_order_status. The current MCP revision is 2026-07-28; ChatGPT also still works with servers on the older initialize-based revisions.
  3. Add authentication. Publish OAuth protected-resource metadata on your MCP server and point it to your authorization server. ChatGPT runs the authorization-code flow with PKCE (S256) and appends a resource parameter that you should copy into the token, usually as aud. It prefers Client ID Metadata Documents where your authorization server supports them, and also supports dynamic client registration or a predefined client. Declare auth per tool with securitySchemes, and return runtime errors carrying _meta["mcp/www_authenticate"] so ChatGPT shows the account-linking prompt.
  4. Allow the right redirect URI. If your authorization server advertises and implements RFC 9207 issuer identification (it returns iss in every authorization response), ChatGPT uses a stable redirect URI. Otherwise it uses a callback-specific one. Copy the exact value from the plugin's management page into your allowlist.
  5. Annotate every tool. Set readOnlyHint, destructiveHint and openWorldHint to explicit true or false values. Review checks that they match real behaviour, and a justification in the form cannot override a wrong value on the server.
  6. Add UI only where it helps. ChatGPT implements the open MCP Apps standard. A tool links to a UI resource (MIME type text/html;profile=mcp-app, usually a ui:// URI) through _meta.ui.resourceUri, and the UI runs in an iframe that talks to ChatGPT over JSON-RPC on postMessage. Declare a narrow content security policy in _meta.ui.csp.
  7. Add extensions. OpenAI's MCP extensions connect your plugin to ChatGPT's sidebar, composer and file viewers through tool metadata.
  8. Add MCP Events if users should be able to say "when X happens in your product, do Y". The next section covers what this takes.
  9. Test, package and submit. Test in ChatGPT Work with a custom connection, bundle the plugin as a ZIP with its plugin.json, skills, icons and MCP configuration, upload it in the portal, fix automated findings, then submit for review. The submission section below has the exact steps.

For enterprise buyers, also request the openid and email scopes and expose a UserInfo endpoint. ChatGPT Enterprise can use the email domain to stop a corporate identity linking your plugin in a personal workspace.

A SaaS product connected to many services
A SaaS product connected to many services

What are MCP events in ChatGPT, and what do they take to build?

MCP Events is a draft, experimental MCP extension that lets ChatGPT subscribe to things that happen in your product, such as a new message or a status change, and start a task when they occur. Your server receives subscriptions and delivers each matching event as a signed webhook. It is not the same as MCP notifications, which only flow over an open connection.

In OpenAI's MCP Events documentation, it works in Work chats on ChatGPT web, in the desktop app with Cloud selected, and with dots. The specification itself is still a design sketch in the MCP project's experimental repository, so expect changes. It is the most backend-heavy part of a plugin.

Your server advertises an events capability (in the server/discover response on the 2026-07-28 revision, or in the initialize response on older ones) and implements three methods on the same authenticated endpoint as your tools:

  • events/list describes each event, its filters and its payload schema.
  • events/subscribe creates or refreshes a subscription. ChatGPT supplies a callback URL and a signing secret with a whsec_ prefix. Make creation idempotent.
  • events/unsubscribe stops it.

Then your server POSTs one signed event per request to the callback when something matches, with a unique eventId and an X-MCP-Subscription-Id header. The production details are what take time:

  • Durable subscription storage that survives restarts, with owner, filters, callback URL, secret and expiry.
  • Authorisation at subscribe time and during the subscription. Check the user may see the event, and stop delivery if their access is revoked.
  • Callback verification with a signed challenge that ChatGPT echoes back before any data is sent.
  • SSRF protection. HTTPS only, resolve and validate the destination address, block private and local addresses, and do not follow redirects.
  • Signing with Standard Webhooks headers, request bodies up to 256 KiB, retries with exponential backoff that keep the same event ID but re-sign each attempt, and no retries on 410 or 413.
  • Idempotent write tools, because events can arrive out of order and an automation may act twice.

OpenAI's integration supports webhook delivery only, not polling or streaming. Its documentation also tells you to treat comments and other user-written text in events as data and never to put instructions for the model in the payload. That matters, because an event-triggered automation is an agent acting on text written by someone else. Our AI agent security checklist covers the controls around that.

How do you submit and publish a ChatGPT plugin as of November 2026?

Verify your identity and domain, upload the plugin ZIP in the OpenAI Platform's Plugins page, connect and scan the MCP server, add reviewer credentials and test cases, then submit for review. After approval you choose when to publish. These are the steps in OpenAI's submission guide as of 13 November 2026:

  1. Set up ownership. Pick the organisation and project that will own the plugin. Organisation owners can submit; other members need the Apps Management Write permission.
  2. Verify identity. Complete individual or business verification in organisation settings. Publishing under an unverified name is rejected.
  3. Verify the domain. Serve the challenge token as plain text at /.well-known/openai-apps-challenge on the MCP hostname or an eligible parent domain.
  4. Upload the ZIP. Choose Upload new or existing plugin, select your verified identity and upload. Keep secrets out of the package. Fix any Metadata & Skills findings.
  5. Connect the MCP server. Under MCPs, connect, authorise and wait for the tool scan, which imports tool names, schemas, annotations, CSP settings and server instructions.
  6. Add review details. Five positive test cases (description, prompt, expected tools, expected behaviour), three negative cases where the plugin should not act, a walkthrough video URL, release notes, and reviewer credentials entered in the dashboard form, never in the ZIP.
  7. Submit and track. Select Submit for review, complete the policy attestations and follow progress under Review status. Only one review can be active per plugin.
  8. Publish. Once approved, open the approved version and select Publish plugin. Approval alone does not list it.
  9. Maintain. OpenAI scans your hosted MCP server daily. Eligible tool changes go live after automated checks; flagged changes are held. Metadata, asset or skill changes need a new ZIP and a new review.

If you are rejected, fix the feedback and submit a corrected ZIP, or reply to the rejection email to appeal.

What does OpenAI's review check before a plugin goes live?

Review checks who you are, whether the plugin works for a reviewer, and whether it follows OpenAI's data and commerce rules. Use this as a readiness checklist, based on the review requirements and plugin guidelines:

  • Verified publisher whose name matches the listing.
  • Domain verification as above.
  • Reviewer access. A dedicated demo account with sample data that works immediately, from outside your network, without MFA, SMS or email codes, or magic links.
  • Passing test cases on every ChatGPT and Codex surface where the plugin is available.
  • Privacy policy that covers the data categories, purposes, recipients, retention and user controls. Review flags undisclosed fields such as internal IDs, telemetry or debug payloads; if a field isn't needed, remove it.
  • Accurate annotations on every tool.
  • No restricted data. Plugins must not collect or process PCI DSS payment data, protected health information, government identifiers, or credentials such as passwords, API keys and MFA codes.
  • Commerce rules. Commerce is limited to physical goods, with checkout on your own domain. A plugin cannot sell subscriptions, tokens or credits, show plans or link to an upgrade checkout, though it may say a feature needs a different plan and link to an informational page.
  • Audience. Suitable for users aged 13 to 17, with no ads.

Three operational details catch teams out. Projects with EU data residency cannot submit plugins with MCP servers; use a global-residency project. The MCP server's origin (scheme, hostname or port) cannot change between versions, so plan your domain before the first submission; OpenAI's submission guide says to contact support if the URL must change. And deleted tools disappear quickly, while new or changed tools go live only after automated checks pass.

Finally, approval is not discovery. Approved plugins must be published manually, and enhanced distribution such as directory placement or proactive suggestions is reserved for plugins OpenAI judges to show strong real-world utility and user satisfaction. You cannot request it.

OpenAI plugin review checklist: verified publisher, domain verification, reviewer demo account, 5 positive and 3 negative test cases, accurate annotations, privacy policy, no restricted data, commerce and audience rules, plus three operational gotchas
Figure 2: OpenAI plugin review checklist: verified publisher, domain verification, reviewer demo account, 5 positive and 3 negative test cases, accurate annotations, privacy policy, no restricted data, commerce and audience rules, plus three operational gotchas

Can a healthcare or fintech SaaS ship a public ChatGPT plugin?

Yes, but not over the regulated data itself. OpenAI's guidelines bar public plugins from processing PHI and PCI-scoped card data. A healthtech or fintech product can still publish a plugin for scheduling, general product information, de-identified analytics or other non-regulated features. For workflows that touch regulated data, a private custom MCP connection inside a customer's workspace is the right route, with the model endpoint, contracts and audit trail scoped like the rest of your compliant system. Our guides to HIPAA-compliant AI agents and the fintech PCI DSS and SOC 2 checklist cover that side.

How long does ChatGPT plugin development take, and how much does it cost?

Plan two to three senior engineer-weeks for a read-only plugin over public data and eight to fourteen for an authenticated plugin with write tools and MCP Events, plus at least one review round. At market rates that is roughly $2.5K to $23K offshore or $8K to $168K at US mid-market rates.

Most of the cost is in authentication, events and review readiness, not the protocol. The table converts effort into money at two published market rate bands: the senior Asia rate of $31–$41 an hour from our 2026 Vietnam rates guide, and the $100–$300 an hour typical of US mid-market firms.

Plugin scopeWhat's includedEffort (senior engineer-weeks)At $31–$41/hAt $100–$300/h
Read-only, public dataMCP server with 5–10 tools, annotations, package, test cases, video, privacy review2–3$2.5K–$4.9K$8K–$36K
Authenticated, read-onlyAbove plus OAuth 2.1 with PKCE, protected-resource metadata, client registration, per-tool auth, demo tenant4–7$5K–$11.5K$16K–$84K
Authenticated with UIAbove plus MCP Apps UI, sidebar or panel extension, content security policy, UI testing6–10$7.4K–$16.4K$24K–$120K
Write tools and MCP EventsAbove plus write scopes, confirmations, idempotency, subscription store, signed webhooks, SSRF guard, abuse testing8–14$9.9K–$23K$32K–$168K

These are planning estimates, not quotes. They assume your API already exposes what the tools need and that the team knows OAuth. If your API does not exist yet, that work comes first, and our AI app development cost guide covers it. Review time is not in the table, because OpenAI says review timelines vary and it does not accept requests for expedited review. Plan for at least one round of feedback.

What did we find when we checked our own MCP servers against the plugin requirements?

Three gaps, which are probably typical for servers built in 2025 or early 2026. BeevR runs three MCP servers: beevr.ai/mcp and ecocheck.ai/mcp (public, read-only) and ecocheck.ai/mcp/account (a customer's own data, OAuth 2.1 with PKCE). On 7 October 2026 we compared them with OpenAI's plugin documentation:

  • No tool annotations (fixed the same day). Both public servers returned tools without readOnlyHint, destructiveHint or openWorldHint. Every tool is read-only, so the fix was one line per server: all three servers now declare readOnlyHint: true, destructiveHint: false, idempotentHint: true and openWorldHint: false on every tool.
  • No MCP Events support. Our servers negotiate the older handshake-based protocol revisions and do not advertise an events capability. OpenAI's current docs accept events on initialize-based servers too, so adding events does not force a protocol upgrade, but the subscription store, signing and SSRF work still apply.
  • The redirect allowlist needs updating. The account server's client registration allows ChatGPT's stable redirect URI. Our authorization-server metadata does not advertise RFC 9207 issuer identification, so under the current docs ChatGPT would use a callback-specific redirect instead. Our authorization response comes straight from Amazon Cognito, so adding iss would mean proxying both the authorize and token steps. We chose the simpler route OpenAI documents: copy the exact redirect from the plugin's management page into the allowlist when we register the plugin.

What already fits: protected-resource metadata, a 401 challenge that names the metadata and scope, S256 PKCE only, public clients, short-lived tokens with revocation, and read-only access enforced in the backend API rather than the MCP layer. If you are starting fresh, building to the current OpenAI docs from day one avoids all three gaps.

FAQ

Is a ChatGPT plugin the same as an MCP server?

Not quite. The MCP server is the part that connects ChatGPT to your product. The plugin is the package you submit and users install, and it can also include skills, UI and metadata. A plugin can even be skills-only with no server.

Do I need a custom UI for my ChatGPT plugin?

No. UI is optional, and many good plugins return structured results only. Add UI where people need to compare, edit or confirm something visually, and keep every tool working without it so agents can use it in the background.

How long does ChatGPT plugin review take?

OpenAI does not publish a timeline and says requests for expedited review cannot be accommodated. Most rejections come from reviewers being unable to log in, failing test cases, undisclosed personal data or wrong annotations, so get those right before you submit.

Will my plugin work with OpenAI dots?

Dots use the supported plugins a user has installed and enabled, so a published plugin is how your product becomes reachable to them. MCP Events also works with dots. Correct annotations matter: OpenAI says a dot's background research is read-only, and follow-up actions need the applicable permissions.

Can the same MCP server work in Claude and ChatGPT?

Yes. Both are MCP clients, and ChatGPT's UI is built on the open MCP Apps standard. Our EcoCheck account server already allows redirect URIs for both clients. Host-specific extras, such as ChatGPT sidebar extensions, need extra metadata but do not break other clients. OpenAI's docs also include a guide for converting a Claude Code plugin or connector into a plugin you can submit.

How BeevR builds ChatGPT plugins

BeevR builds production MCP servers and the plugins on top of them, with OAuth, audit and review readiness designed in. We start from your three to five use cases, write the review test cases first, and ship a working demo every Friday. Pricing is fixed per phase, from $10K per phase, and you own the code from day one. See our AI agent development work.

Book a 30-minute call and tell us what you want ChatGPT users to do with your product.