Custom GPTs
People in a Business, Enterprise or Edu workspace who have a task colleagues keep asking them to do, and want to hand it to a GPT instead
Prerequisites: Module 303 (ChatGPT Projects) or equivalent experience. Module 304 (Connectors and Apps) helps.
Last verified 2026-09-17
Learning objectives
By the end, participants can:
- State the current position on GPTs (custom GPTs are being retired in favour of Plugins: new GPT creation is withdrawn from Free, Go, Plus and Pro, the migration flow is targeted for 17 September 2026, and Enterprise retirement is planned for 11 December 2026) and what it means for their plan
- Configure a GPT: name, description, starters, instructions, knowledge, capabilities, apps or actions
- Separate behaviour (instructions) from reference material (knowledge, up to 20 files, each up to 512 MB)
- Test a GPT in preview against realistic inputs and fix instructions before adding features
- Share a GPT at the right level and permission, and know what a workspace admin controls
- Build so the workflow ports to a plugin
What's in this module
6 lessons · 5 exercises · ~7 min reading
Starts withWhere GPTs stand in September 2026~1 minExecutive summary
You will learn how to build, test and share a custom GPT for a repeated task: instructions, knowledge files, capabilities, apps or actions, and the sharing levels a workspace allows. You will also learn where GPTs are going, because OpenAI is retiring them in favour of plugins, and how to build so the move is cheap. End state is one working GPT for a real repeated task, tested against three inputs, shared with the people who need it, and a written note of what will move to a plugin. After this module you can turn a task you explain weekly into something a colleague runs alone, decide what belongs in instructions versus knowledge, and choose between a GPT, a shared Project and a plugin.
Sources
- OpenAI Help Center, "GPTs in ChatGPT", what a GPT is, current plan availability, the retirement and migration timeline, and the GPT-versus-Project comparison: https://help.openai.com/en/articles/8554407-gpts-in-chatgpt
- OpenAI Help Center, "Creating and editing GPTs", the configuration fields, knowledge limits, testing and versioning: https://help.openai.com/en/articles/8554397-creating-and-editing-gpts
- OpenAI Help Center, "Sharing and publishing GPTs", sharing levels, permission levels and GPT Store requirements: https://help.openai.com/en/articles/8798878-sharing-and-publishing-gpts
- OpenAI Help Center, "Plugins in ChatGPT and Codex", what a plugin is and how workspaces manage them, the destination for migrated GPT workflows: https://help.openai.com/en/articles/20001256-plugins-in-chatgpt-and-codex
- OpenAI Help Center, "Projects in ChatGPT", the shared-Project alternative for team work: https://help.openai.com/en/articles/10169521-projects-in-chatgpt
Prefer the live room?
This module also runs inside our in-person bootcamps and workshops in Copenhagen.
More in Track C
Claude Skills
How skills turn your repeated tasks into reusable, auto-triggering workflows.
Claude Connectors
How to connect Claude to the tools where your work already lives, and the restraint rules that keep it cheap and safe.
Cowork + Projects
How Projects give Cowork persistent memory, standing instructions, and scheduled tasks.