This project helps analyze your ChatGPT usage data, particularly in light of usage limits (e.g., for "o3" inference models). By processing your exported conversation data, you can get insights into your model utilization patterns.
(chatgpt-usage) ⋊> ~/chatgpt-usage on main ◦ uv run main2.py 19:20:01
Calculating model usage for the period:
Start: 2025-05-12 00:00:00 UTC (Timestamp: 1747008000.0)
End: 2025-05-18 23:59:59 UTC (Timestamp: 1747612799.999999)
------------------------------
Model Usage Counts:
gpt-4o: 52
o3: 53
o4-mini-high: 114
research: 3
o4-mini: 6-
Export Your Data:
- Go to your ChatGPT account settings.
- Request an export of your data.
- You will receive an email with a download link.
-
Prepare
conversations.json:- Download the exported data (usually a
.zipfile). - Unzip the file. You should find a
conversations.jsonfile inside. - Place this
conversations.jsonfile in the same directory as the Python scripts (main.py,main2.py).
- Download the exported data (usually a
-
Run Analysis Scripts:
-
To analyze all your conversation data:
python main.py
The results will be logged to
main.log. -
To analyze data from the last 7 days:
python main2.py
The results will be printed to the console.
-
main.py: Processes the entireconversations.jsonfile and logs detailed model usage statistics, both overall and per conversation.main2.py: Processesconversations.jsonbut filters for messages from the last seven days, providing a summary of recent model usage.