OpenAI launches the ability for paying ChatGPT users to invoke GPTs directly in chats by typing “@” and selecting a GPT from a list
Context & Ripple Effects
OpenAI had already positioned custom GPTs as no-code creations for Plus and Enterprise users, with a planned store and monetization path in its initial GPT rollout. This update makes those GPTs easier to reach from the core conversation rather than treating them as separate destinations.
It also extends an earlier paid-tier strategy: OpenAI had been considering ways to monetize ChatGPT well before broadening the product experience for free users with voice.
First-order effects
- Paying ChatGPT users can switch to a relevant custom GPT inside an existing chat through an @ mention, reducing the friction of finding and opening a separate GPT.
- GPT creators gain a more direct in-product route to usage, while paid subscribers receive a clearer functional distinction from users without access.
Second-order effects
- Usage may concentrate around GPTs that are easy to identify and invoke in a live conversation, raising the value of naming, discovery, and task-specific positioning for creators.
- The change puts pressure on competing AI assistants to make specialized tools available within the main chat flow rather than behind separate apps or menus.
Third-order effects
- If in-chat invocation becomes a standard interface pattern, the assistant can evolve from a single model experience into a work surface that routes users among specialized capabilities.
- That would make curation, discoverability, and eventually commercial terms for third-party assistants more consequential than simply publishing a custom prompt.
The trend: This is one step toward chat interfaces becoming an assistant operating layer where specialized AI tools are called from the same conversation.