Tracks and displays ChatGPT message usage based on model limits, with a toggle button to reopen the info panel. Adds support for gpt-4 model, makes model usage collapsible, and persists collapse state between page reloads.
The ChatGPT Message Tracker is a userscript designed to help you monitor and manage your message usage while interacting with ChatGPT. This tool tracks the number of messages you send based on the model you use and provides real-time statistics on your remaining message limits.
Network Monitoring: Automatically intercepts network requests sent to the ChatGPT API, specifically targeting the conversation endpoint.
Model-Specific Tracking: Keeps track of usage limits for various models, including:
Rolling Windows: Utilizes rolling time windows to accurately calculate remaining messages based on the specified limits.
Persistent Storage: Saves your message usage data and the collapse state of the tracker in localStorage, ensuring that your statistics and preferences persist across browser sessions and restarts.
User-Friendly Interface: Displays a floating panel with your current message usage and remaining limits, with an option to close and reopen the tracker via a toggle button. The visibility of each model's usage details can be collapsed or expanded, with the state remembered between page reloads.
I'm very happy to accept issues or pull requests on the GitHub repository.
This script is distributed under the MIT License.
gpt-4 model with a limit of 40 messages per 3 hours, which counts towards the gpt-4o usage cap.localStorage.