iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
You can create or edit a custom GPT on ChatGPT’s web interface if your account and workspace permissions allow it. OpenAI’s current guidance says personal Free, Go, Plus, and Pro accounts cannot create or publish new GPTs; eligible Business, Enterprise, and Edu workspace members may be able to, depending on workspace settings. Start by confirming eligibility, then give the GPT one clear job, put behavioral rules in Instructions and reference material in Knowledge, and test it in Preview before sharing or updating it.
Check whether your account can create or edit GPTs
GPTs are configured versions of ChatGPT for a specific purpose, built with instructions, knowledge, and selected capabilities. Creation and editing are done on the web; mobile apps support using GPTs, not creating them. OpenAI’s current creation and editing guidance says personal Free, Go, Plus, and Pro accounts cannot create or publish new GPTs. Eligible Business, Enterprise, and Edu workspace members may create them, but workspace settings and permissions still apply. Editing an existing GPT is also subject to plan and permission eligibility.
Before investing in a build, check your account’s GPT editor access and ask a workspace administrator about relevant permissions. Availability and sharing options can vary by account, workspace configuration, and region.
Define a focused job before opening the editor
A useful GPT has a specific audience and outcome. For example, “help new employees find the right onboarding procedure” is easier to configure and test than “answer any question about the company.” Decide what it should do, what it should not do, and what information it needs to perform the job.
#1 Best Overall
Open the GPT editor on the web and choose either the conversational builder, which helps draft a configuration through prompts, or direct configuration, which gives you explicit control over its fields. Give it a clear name and description. Add realistic conversation starters that show the kinds of requests it handles; these help users understand how to begin, but they do not replace instructions.
Put behavior in Instructions and reference material in Knowledge
| Configuration area | What belongs there | Example |
|---|---|---|
| Instructions | Behavior, tone, decision rules, boundaries, and workflow | Ask for the user’s department before recommending an internal procedure; answer in concise steps. |
| Knowledge | Reference documents the GPT should draw on | An employee handbook, product guide, or internal process document. |
Write instructions as explicit, readable steps. State the intended audience and task, specify how to handle uncertainty or missing information, and include examples when a preferred response format matters. Keep the reference files focused on useful source material rather than asking them to act as behavior rules.
OpenAI’s current guide allows up to 20 Knowledge files per GPT, each up to 512 MB. Accepted formats vary by model and may depend on Code Interpreter & Data Analysis being enabled. OpenAI recommends clear, text-forward files; documents with complex layouts may be harder for a GPT to use reliably.
Windows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallOutdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchChoose capabilities only when they support the job
Built-in capability options listed by OpenAI include web search, image generation, Canvas, Code Interpreter & Data Analysis, and connected apps. Which options are available depends on the account, workspace setup, and region. Enable only what the GPT needs: adding tools can introduce extra behavior to test and may not address a problem caused by unclear instructions.
A recommended model can guide users toward a suitable starting point, but it does not prevent them from switching to another model when one is available to them.
Use an action for a custom API connection
Actions let a GPT connect to an external API. In the GPT editor, configure the service’s API details, select an authentication method—none, API key, or OAuth—and provide an OpenAPI schema in JSON or YAML. Test the action in Preview before relying on it. OpenAI’s actions configuration guide explains the setup and requirements.
- A GPT can use connected apps or actions, but not both at the same time.
- Workspace domain allowlists can prevent an action from executing.
- Public GPTs that use actions need a valid privacy-policy URL.
- Users may be prompted to approve action runs.
Test the GPT in Preview, then update deliberately
Use Preview with realistic prompts before sharing or applying changes. Check whether the GPT follows its instructions, uses Knowledge appropriately, responds in the intended tone, and handles capabilities or actions as expected. Include ordinary requests as well as edge cases, such as a question the reference files do not answer.
OpenAI advises improving instructions and adding examples before adding more tools: “Before adding more tools, tighten instructions and add examples. This often fixes issues faster than adding features.” (OpenAI Help Center, Creating and editing GPTs.)
Best Value
Edits save as a draft. When you are ready to apply changes to an existing GPT, select Update. Version history lets you review or restore earlier versions. If you restore an older version that uses actions, you may need to configure its authentication again.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Choose sharing settings and check time-sensitive notices
Sharing choices depend on account and workspace settings. Before sharing, verify who can access the GPT, whether workspace permissions allow the intended audience, and whether publication requirements apply. OpenAI’s sharing and publishing guidance covers the available options and requirements.
OpenAI’s current creation and editing article displays a planned retirement and migration notice for affected Enterprise workspaces: retirement is planned for December 11, 2026, with migration targeted for September 17, 2026. This is time-sensitive and account-specific; verify the live notice and your workspace’s instructions before building a long-term workflow around an affected GPT. The article describes migration to “Plugins,” so confirm the terminology and applicable plan details in the current notice rather than assuming it applies to every account.
The Tool Desk
Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
