#With MCP embraced by AI dev community, are CustomGPTs a thing of the past?

1 messages · Page 1 of 1 (latest)

warm ember
#

With MCP embraced by the entire AI community (developers, LLM service providers, IT biz, software vendors, customers), are CustomGPTs a thing of the past? Seems to me they serve largely the same purpose, with CustomCPTs being an OpenAI thing, Custom GPTs having a narrower scope and being closely tied to the OpenAPI spec for tool support.

So MCP is more versatile, more comprehensive and more broadly suppported. Has OpenAI made some remarks about the strategic role and future support of Custom GPTs?

midnight knotBOT
#

Hi @warm ember. I'm kapa.ai, your friendly support bot. I will try to answer your question by searching through the knowledge base. :wave:

#

There is no information in the provided knowledge sources indicating that OpenAI considers Custom GPTs to be obsolete or that they are being replaced by MCP (Model Context Protocol). In fact, the latest release notes and help articles show that Custom GPTs continue to receive new features and expanded model support, including the ability for creators to select from all available ChatGPT models and set recommended models for their GPTs. GPTs with custom actions currently support GPT-4o and 4.1, while those without actions can use any available model. This functionality is being rolled out across Plus, Pro, Team, Enterprise, and Edu plans, and there is no mention of deprecation or reduced support for Custom GPTsExpanded Model Support for Custom GPTs.

MCP is described as a protocol for building custom connectors to proprietary systems, primarily for use in "deep research" scenarios within ChatGPT for Team, Enterprise, and Edu customers. MCP enables integration with a broader range of internal and external tools, but it is positioned as a way to extend ChatGPT’s capabilities, not as a replacement for Custom GPTs. MCP connectors are available only in deep research and require admin setup, whereas Custom GPTs are designed for a broader audience, including non-developers, and can be created and shared easily within ChatGPTCustom connectors via Model Context Protocol (Beta).

#

OpenAI has not made any public remarks in the provided sources about phasing out Custom GPTs or shifting strategic focus away from them. Instead, both Custom GPTs and MCP-based connectors are being developed in parallel, serving different user needs: Custom GPTs for accessible, no-code customization and sharing, and MCP for advanced, developer-driven integrations.

In summary, based on the current documentation, Custom GPTs remain a core, actively supported feature of the OpenAI ecosystem, and there is no indication that they are being replaced by MCP or are becoming obsolete.

Note: you can ask me a follow up question by @ mentioning me again :speech_balloon: