Contributors:
Amol Ghemud Published: January 11, 2024
Summary
The GPT Store is OpenAI’s public library of custom GPTs, opened on January 10, 2024 and browsed through Explore GPTs. As of September 2026, all signed-in ChatGPT users can use GPTs, but only Business, Enterprise and Edu workspaces can create or publish new ones. OpenAI plans to retire custom GPTs across ChatGPT plans and move workflows to plugins, with retirement scheduled for December 11, 2026 in Enterprise workspaces and other plans expected to follow.
In This Article
Share On:
Imagine the most relevant GPTs served to you from a pool of more than 3 million. That was the pitch when OpenAI opened the GPT Store on January 10, 2024: a public library where ChatGPT Plus, Team and Enterprise users could find and share custom versions of ChatGPT.
Since that launch, Free users gained GPT access, personal plans lost GPT creation and OpenAI now plans to retire custom GPTs. This explainer is updated for September 2026, with launch details dated.
What Is the GPT Store? (Quick Answer for 2026)
The GPT Store is OpenAI’s public library of custom GPTs, which you browse through Explore GPTs in ChatGPT. As of September 2026, any signed-in user can chat with existing GPTs, but only Business, Enterprise and Edu workspaces can create or publish new ones, and OpenAI plans to retire custom GPTs.
OpenAI’s custom GPT retirement and migration FAQ says the transition affects all ChatGPT plans. Enterprise workspaces retire on December 11, 2026, other plans are expected to follow, and dates are subject to change.
How to Explore GPTs in ChatGPT
Open Explore GPTs in ChatGPT (chatgpt.com/gpts), pick a GPT and sign in to start a chat. You can also use a direct link or GPTs shared in your workspace.
Open Explore GPTs at chatgpt.com/gpts
OpenAI’s GPTs in ChatGPT help article says GPTs are available to all ChatGPT users. Public GPT pages may be visible without an account, but you need to sign in to start a conversation.
Bring a GPT into any chat with @
On ChatGPT web, type @ in a conversation to bring in a GPT. This doesn’t work in the iOS or Android apps. Builders can’t see your individual conversations with their GPTs.
What OpenAI Announced at the GPT Store Launch (January 2024)
At launch, the store was open to ChatGPT Plus, Team and Enterprise users, showed trending GPTs by category and promised a revenue program for US builders.
A delayed opening and 3 million GPTs
OpenAI introduced GPTs on November 6, 2023, saying no coding was required and the store would launch later that month. It opened on January 10, 2024, when OpenAI’s launch announcement said users had created more than 3 million custom versions of ChatGPT.
Discovery and builder rules at launch
The launch post showed trending GPTs in DALL·E, writing, research, programming, education and lifestyle, plus new featured GPTs every week. To list a GPT, builders saved it for Everyone and verified a builder profile with their name or a verified website.
The builder revenue promise
OpenAI said it would launch a GPT builder revenue program in Q1 2024, starting with US builders paid based on user engagement with their GPTs.
OpenAI’s creating and editing GPTs article says personal accounts, including Free, Go, Plus and Pro, can’t create or publish new GPTs. GPTs you built earlier remain available to use, and editing depends on your current plan and permissions.
What happened to the builder revenue program?
We couldn’t find a current OpenAI page describing an active GPT builder revenue program. OpenAI’s old “Monetizing your GPT” help address now opens the sharing and publishing article, which doesn’t mention builder payments, so don’t build a GPT expecting payouts.
Who Can Build and Publish Custom GPTs in 2026?
Only Business, Enterprise and Edu workspaces can create, edit and publish new GPTs, and only when workspace settings and role permissions allow it.
Confirm the GPT meets policy and product requirements.
Builder profile verification and blocked GPTs
Your builder profile can show your name, a website on a verified domain and supported verified social profiles. GPTs pass automatic checks before public sharing. If one is blocked, update and retry, appeal or keep it private.
What a custom GPT can do
Builders can add web search, image generation, canvas, data analysis and up to 20 knowledge files of up to 512 MB each.
Custom GPTs Are Being Retired: Dates and What Replaces Them
OpenAI plans to retire custom GPTs across ChatGPT plans and recommends moving workflows to plugins. Enterprise workspaces have a scheduled retirement date of December 11, 2026.
The planned timeline
OpenAI’s ChatGPT release notes entry of September 11, 2026 announced the plan. The FAQ’s Enterprise dates, subject to change:
September 11, 2026: admin notice.
September 17, 2026: migration experience and user banner (targeted, and may not reach every account at once).
September 25, 2026: creation of new custom GPTs ends, while existing GPTs stay usable.
December 11, 2026: retirement, when custom GPTs stop running.
What plugins are
Plugins in ChatGPT and Codex package reusable instructions, connected tools and other capabilities. OpenAI says the plugin directory is available across ChatGPT plans. When a GPT migrates, its instructions become a skill in the new plugin and the original GPT turns read-only.
What to do if you rely on GPTs
Save each GPT’s instructions and knowledge files now. Custom actions don’t transfer, so plan to rebuild them with connectors or custom MCP servers. For public GPTs from other builders, what matters is where each was created, and access doesn’t guarantee access to the replacement plugin.
Browsing the GPT Library by Category (With upGrowth’s GPTs)
The 6 launch categories are still a handy way to think about the GPT library. Here are upGrowth’s own custom GPTs in each.
Trending News Ocean curates the latest news by topic. Digital Fitness Coach offers diet plans, gym tips and recipes. Like every custom GPT, these fall under OpenAI’s retirement plan.
Yes, for now. As of September 2026, OpenAI says all signed-in ChatGPT users can use GPTs through Explore GPTs. OpenAI plans to retire custom GPTs across plans: Enterprise workspaces retire on December 11, 2026, other plans are expected to follow, and dates are subject to change.
How do I explore GPTs in ChatGPT?
Open Explore GPTs in ChatGPT, or go to chatgpt.com/gpts, and pick a GPT. Public GPT pages may be visible without an account, but you must sign in to chat. On ChatGPT web, type @ in a conversation to bring a GPT in. That doesn’t work in the mobile apps.
Can I still create a custom GPT on ChatGPT Plus?
No. OpenAI says personal accounts, including Free, Go, Plus and Pro, can’t create or publish new GPTs. Only Business, Enterprise and Edu workspaces can, when workspace settings and permissions allow it. GPTs you built earlier on a personal plan remain available to use.
Does OpenAI pay GPT builders?
In January 2024, OpenAI said it would launch a GPT builder revenue program in Q1 2024, paying US builders based on user engagement. As of September 2026, we couldn’t find a current OpenAI page describing an active program, and its old monetization help address now opens an article with no payment details.
What is replacing custom GPTs?
OpenAI recommends moving custom GPT workflows to plugins, which package reusable instructions, connected tools and other capabilities. A migrated GPT’s instructions become a plugin skill. The migration experience was targeted for September 17, 2026, but may not reach every account at once. Custom actions don’t transfer.
Your Next Move: Plan for Life After Custom GPTs
If your team runs daily work through custom GPTs, start the move to plugins now, not in December.
The GPT Store was initially introduced as a public, app-store-like marketplace for user-generated AI applications, aiming to cultivate a broad ecosystem of custom tools. By September 2026, its function has pivoted dramatically to a legacy library of existing GPTs with a confirmed retirement plan, signaling OpenAI's strategic move away from a public creator model. This evolution redefines its value from an innovation hub to a temporary feature for specific account types. The initial launch highlighted trending GPTs and promised a builder revenue program, positioning it as a vibrant community. Now, with creation restricted to Business, Enterprise, and Edu workspaces and a sunset date announced for December 11, 2026 for some plans, the focus has clearly shifted toward controlled, enterprise-grade AI solutions over an open creator economy. This change underscores a larger trend in managing AI development. Examining this transition helps clarify where the future of custom AI tools on the platform is headed.
Custom GPTs initially served as a strategic pillar for OpenAI to drive user engagement and demonstrate the platform's versatility, crowdsourcing innovation from a base that created over 3 million GPTs by launch. Their discontinuation marks a significant pivot from a strategy of decentralized, user-led creation to one of centralized, high-value enterprise services. This matters because it changes how businesses and creators can build on the ChatGPT platform, pushing them toward more formal API integrations instead of simple custom interfaces. The retirement, starting with Enterprise workspaces on December 11, 2026, indicates that the open-ended experiment in user-generated AI is concluding in favor of more structured, predictable, and monetizable offerings. This shift away from the initial community-focused model impacts anyone who invested time in building a presence on the GPT Store. You must now re-evaluate your long-term AI strategy in light of this platform direction.
The primary difference in the GPT experience between a ChatGPT Plus user and a ChatGPT Enterprise user is the total loss of creation ability for personal plans. While both can still use existing GPTs, only users in Business, Enterprise, and Edu workspaces can create or publish new ones, fundamentally splitting the user base into consumers and creators. This creates a clear delineation between personal-use consumption and professional-use development within the OpenAI ecosystem. An Enterprise user can build and share GPTs within their workspace's private section, while a Plus user is limited to accessing public GPTs through 'Explore GPTs' or direct links. Furthermore, Enterprise workspaces have a set retirement date of December 11, 2026, forcing them to plan a migration strategy, whereas other plans are still awaiting a final timeline. This distinction is critical for teams evaluating which OpenAI plan aligns with their long-term objectives for custom AI development.
The rapid creation of over 3 million GPTs demonstrated massive user interest, but OpenAI's decision to restrict creation and plan a retirement signals a strategic retreat from managing a massive, open ecosystem. This pivot suggests that the costs and complexities of moderating a public AI store outweighed the benefits of user-generated innovation. The platform's direction now favors controlled, high-quality experiences, particularly for enterprise clients. Initial enthusiasm, reflected in the 3 million figure, did not translate into a sustainable model for OpenAI, especially with the promised revenue-sharing program never materializing. The focus has shifted to structured plans like ChatGPT Business, where GPTs are internal tools rather than public products. This move indicates that while user-generated content is a powerful growth engine, AI platforms may find more value in curated, professional-grade applications. Understanding this shift is vital for predicting future opportunities on large AI platforms.
The evolution from a curated storefront with categories like DALL·E and writing to a search-based 'Explore GPTs' interface and in-chat '@' mentions reveals a shift in product philosophy. OpenAI has moved from promoting passive discovery to enabling intentional, context-specific retrieval. This change suggests the initial app-store model, with trending and featured sections, was not effective for navigating the 3 million available GPTs. The current system prioritizes utility over browsing. It assumes you either know what you are looking for or want to pull a specific tool into an ongoing conversation. This approach reduces the burden on OpenAI to curate and rank a vast library, instead putting the onus on users to find and use tools directly. This change reflects a broader product trend where functionality is embedded directly into the primary workflow rather than housed in a separate, destination-style marketplace. This insight is key for understanding how to design and surface AI tools effectively.
The unfulfilled promise of a builder revenue program indicates significant challenges in creating viable monetization models for third-party AI creators. It suggests that OpenAI found it difficult to measure and reward 'user engagement' in a fair and scalable way across millions of GPTs, a problem that often plagues user-generated content platforms. This situation serves as a cautionary tale about the economic instability of building on nascent, rapidly changing AI ecosystems. The initial announcement in January 2024 generated excitement, but its quiet disappearance from help articles by September 2026 implies that technical or economic hurdles were too high. This outcome may push serious developers away from platform-specific, 'no-code' solutions and toward more stable, API-based business models where they control their own monetization. For creators, this history underscores the importance of platform independence when developing commercial AI applications.
To integrate a custom GPT into a workflow, you can call it directly within any conversation on the ChatGPT web interface, a feature that streamlines access to specialized tools. This method avoids the need to navigate away from your current task to the 'Explore GPTs' page, making your process more efficient. Here is the process for a ChatGPT Business user:
Begin a new conversation or open an existing one in the ChatGPT web application.
In the message input box, type the '@' symbol. A menu will appear listing recently used and popular GPTs.
Start typing the name of the desired GPT. The list will filter in real time to help you find the correct one.
Select the GPT from the list. Its name will now appear in the input box, ready to receive your prompt.
This approach transforms GPTs from standalone applications into on-demand specialists you can consult mid-task. Note this feature is exclusive to the web version and not available in the mobile apps as of September 2026. Mastering this workflow is essential for maximizing team productivity.
The planned retirement of custom GPTs means that businesses must prepare for significant workflow disruptions and begin developing a migration plan immediately. The key implication is that any reliance on the 'no-code' GPT-building interface is temporary, forcing a shift toward more durable, code-based solutions using the OpenAI API. Strategic preparation involves auditing existing GPTs, identifying critical functions, and re-platforming them before the deadline. Businesses should act now to avoid a last-minute scramble. Key steps for preparation include:
Audit Usage: Identify which custom GPTs are most critical to your team's daily operations.
Preserve Prompts: Document the core instructions and knowledge files used in each essential GPT.
Develop an API-Based Alternative: Begin building replacement applications using the official API, which offers more stability and control.
The December 11, 2026 deadline is a clear signal from OpenAI to transition toward more robust integrations. Proactive migration will ensure business continuity and reduce dependence on features with a limited lifespan.
OpenAI's decision to separate access from creation rights clarified its strategic direction by segmenting users into consumers and professional builders. Initially, granting GPT access to all users in May 2024 democratized usage, but later restricting creation to Business, Enterprise, and Edu workspaces solved the problem of managing a chaotic, low-quality public marketplace. This strategic bifurcation allows OpenAI to encourage broad adoption while channeling serious development through paid, manageable tiers. This approach addresses the core challenge of balancing open access with ecosystem quality and support costs. By preventing new GPT creation on personal plans (Free, Go, Plus, and Pro), OpenAI reduces the moderation burden and nudges serious creators and businesses toward its premium offerings. This move provides a clear answer to who the platform is for: consumption is for everyone, but creation is for professional teams. This clarity helps users select the right plan for their needs.
The primary operational risk is a complete loss of control and access to critical business tools, as GPTs are tied to the personal accounts that created them. With creation now disabled for personal plans and employees potentially leaving the company, these tools become uneditable, unmanageable, and at risk of deletion. This creates a significant business continuity problem rooted in the use of non-enterprise software for business purposes. The solution requires a formal migration and centralization strategy. Companies facing this issue must:
Conduct an Immediate Audit: Identify all business-critical GPTs running on personal employee accounts.
Centralize on a Business Plan: Upgrade to a ChatGPT Business or Enterprise plan, which provides a private workspace for team-owned GPTs.
Recreate and Redeploy: Systematically recreate the essential GPTs within the new, company-controlled workspace.
This proactive approach solves the ownership problem and aligns the company with OpenAI's intended use case for professional teams, ensuring tools remain corporate assets.
The lifecycle of custom GPTs, from introduction in November 2023 to a planned retirement starting in 2026, suggests that feature lifecycles on major AI platforms are accelerating dramatically. This condensed timeline indicates that platforms like OpenAI are comfortable launching large-scale experiments and quickly deprecating them if they do not align with long-term strategic goals, such as enterprise focus and profitability. This rapid iteration model requires users and developers to adopt a more agile and less dependent mindset when building on new features. Key phases in this lifecycle included:
November 2023: Introduction of GPTs.
January 2024: GPT Store launch with 3 million GPTs.
May 2024: Access expanded to free users.
2025-2026: Creation restricted and retirement announced.
This trajectory serves as a strong indicator that businesses should favor stable, documented APIs over experimental, no-code interfaces for any critical applications. The future will likely see more such rapid product cycles from OpenAI and its competitors.
With GPT creation now restricted to workspace plans, the officially recommended path for personal account users is to transition to using the OpenAI API. This approach offers far more power, stability, and control than the custom GPT interface ever did, though it does require some coding knowledge. This shift effectively moves custom AI development from a 'no-code' hobbyist environment to a more professional, developer-centric one. While you can continue using existing GPTs created before the rules changed, you cannot edit or create new ones. The API allows you to replicate and even enhance the functionality of a custom GPT by defining the model's behavior, providing it with specific instructions, and connecting it to external tools or data sources. For those without coding skills, this change may mean upgrading to a ChatGPT Business plan if they wish to continue building within a guided interface. This bifurcation is central to OpenAI's current strategy.
Amol has helped catalyse business growth with his strategic & data-driven methodologies. With a decade of experience in the field of marketing, he has donned multiple hats, from channel optimization, data analytics and creative brand positioning to growth engineering and sales.
