GPT Tool Use Benchmark Chart
T²-bench: tool use GPT-5 OpenAI o3 GPT-4.1 Accuracy (%) 97 58 34 81 80 74 63 65 56 Telecom Retail Airline
A bar chart compares GPT-5, OpenAI o3, and GPT-4.1 on tool-use accuracy across telecom, retail, and airline tasks. The joke is the stark performance gap, especially GPT-5’s 97% score in telecom, making it a reaction to debates about which AI model is best at using tools.
- @dev_meme / 7000 2025-08-07
- Type
- image
- Size
- 800×441 · 20.4 KiB
- Mood
- analytical
- Language
- en
- Added