ChatGPT PersonalizationBuilder · schema v2
Reusable ChatGPT personalization profiles

ChatGPT Personalization Builder

Create, validate, and export a reusable personalization profile. Start with General for a balanced English/Indonesian setup, or Blank for no behavioral defaults.

No sign-in No runtime dependencies Edits stay in your browser
01Choose a preset
02Customize your profile
03Validate and apply
Starting point
Validation target
Loading profile…

Profile

Edit profile

Change structured profile fields. The rendered output updates automatically.

Advanced JSON editorDirect schema-level control

Edit the complete v2 document directly. Applying JSON rebuilds the visual editor and preview.

Rendered output

Apply to ChatGPT

Apply the product settings manually, then copy the generated text into the matching Personalization fields.

Recommended order: Match Product settings first, then paste Occupation, More about you, and Custom Instructions into the corresponding ChatGPT Personalization fields.

01

Product settings

Apply manually

          
02
Occupation
Paste → Occupation

          
03
More about you
Paste → More about you

          
04
Custom Instructions
Paste → Custom Instructions

        


        

          
Where to put these fields
Open ChatGPT → Settings → Personalization. Match the product settings above, then paste Occupation, More about you, and Custom Instructions into their corresponding fields. Product labels can change over time; use the closest current field.

          
Need version control?
Download the profile JSON or use the repository CLI to lint, render, compare, and evaluate profiles locally.

        

      

    


    

      
      

        
Profile roles

        
Reusable presets stay separate from personal profiles.

      

      

        
General and the role presets are reusable starting points. Blank contains no behavioral defaults. Maintainer and operational profiles are public examples or account targets, not presets for other users.

        
Structural validation checks configuration and character limits. Behavioral changes still need representative evaluation.