Model Truth for ChatGPT per Dgraciano

You picked a model. Did it answer? Model Truth puts a small label under every ChatGPT
answer naming the model that wrote it, read from the conversation itself.

If an answer came from a different model than most of the chat, the label turns amber and
says so, for example "GPT-4o mini (not GPT-4o)". That covers silent switches when you hit a
limit, and your own switches too.

• One small label per answer, nothing else on the page changes
• Works on old chats as well as new ones
• Usage counter: how many answers each model wrote for you in the last 3 hours, with the
count on the toolbar icon

ChatGPT doesn't show how much you have left until you hit a limit. Model Truth counts as you
go, so you can see it coming.

Pro ($4.99 once):
• 24-hour and 7-day history per model
• Caps: enter your plan's limit for a model (per 3 hours, day or week); the toolbar badge
turns amber at 80% and red at the cap

Free: the labels and the 3-hour count, no account needed.

Private by design: runs only on chatgpt.com and reads the conversation with your own session
in your browser. The usage counter keeps only the model name and time of each new answer, in
your browser, for 7 days. Nothing is sent to us or anyone else. No analytics.

Not affiliated with OpenAI.

Watch a short demo with sample data.