Feature request
I would like to be able to use an existing private or workspace Custom GPT directly inside a ChatGPT Project.
At the moment, Custom GPTs and Projects seem to work as separate systems. A Custom GPT can already contain carefully configured instructions, knowledge files, capabilities, and sometimes actions. A Project, on the other hand, provides a shared workspace with project-specific chats, files, and project instructions.
The missing link is that I cannot simply apply or select one of my existing Custom GPTs as the assistant for a Project. Instead, I have to manually recreate the GPT inside the Project by copying instructions, uploading the same knowledge files, and trying to reproduce the same behavior. This is redundant, error-prone, and undermines the value of reusable Custom GPTs.
Proposed solution
Please add a way to use a Custom GPT inside a ChatGPT Project.
Possible implementation options:
-
Link GPT to Project
The Project uses the selected Custom GPT live. Changes to the GPT would automatically affect the Project. -
Import GPT configuration into Project
The Project copies the GPT’s instructions, knowledge configuration, capability settings, and other relevant setup once. -
Select Project Assistant
A Project could have a setting such as “Assistant” or “GPT for this Project,” where users can choose from their own private GPTs or workspace GPTs.
Instruction conflict handling
There should be a clear rule for how GPT instructions and Project instructions interact.
For example:
-
GPT instructions define the assistant’s general working style and domain behavior.
-
Project instructions define project-specific context and constraints.
-
If there is a conflict, the UI should clearly indicate whether Project instructions override GPT instructions or whether both are merged.
Ideally, users should be able to see which behavior comes from the GPT and which comes from the Project.
Why this matters
A Custom GPT defines how work should be done.
A Project defines what the work is about.
Combining both would make Projects much more powerful and reduce duplicated setup work.
Benefits
-
Avoids manually recreating GPTs in every Project
-
Makes Custom GPTs more reusable
-
Keeps behavior consistent across different Projects
-
Reduces errors from copied or outdated instructions
-
Saves time when working with specialized GPTs
-
Makes Projects more useful for professional and long-running workflows
-
Supports teams that already use workspace GPTs as standardized assistants
Example use case
I have a private Custom GPT configured as a specialized assistant for a particular workflow. It has specific instructions, terminology, and reference files. I now want to use that same assistant inside a Project that contains related chats and files.
Currently, I have to rebuild the GPT manually inside the Project. Instead, I would like to select that GPT as the Project’s assistant, so the Project can use the same behavior and knowledge without duplication.