It's the end of the year again, and various platforms have started to release their annual summary reports. Recently, ChatGPT also released its annual report, but upon reviewing it, it feels not quite accurate -_-

Additionally, since I've used quite a few large model platforms this year, I decided to take matters into my own hands, spending some time extracting the questions I frequently asked on various platforms and analyzing them. Finally, I compiled a personal annual usage report covering all platforms.

The data sources used for this analysis include: ChatGPT, Kimi, Claude, Claude Code, DeepSeek, and Gemini, totaling 5646 raw question records collected. After completing screening and preprocessing, 5621 valid questions with 8454 pieces of analysis data were obtained. Through simple semantic analysis and clustering, the following results were achieved. The usage report is rendered with Excalidraw; if you are an RSS reader, please click the original text to view the report. Sending hearts~

Finally, today is the last day of 2025, so let the giraffe wish you a happy 2026~