ChatGPT draws itself, and then we take its words away
ChatGPT · 3 attempts
Creative & Weird benchmark
The prompts nobody benchmarks.
Odd, open-ended tasks that reward imagination — and expose how differently these models interpret an instruction.
1 tests completed
The record
Leaderboards How the models compare in Creative & Weird
| Model | Wins | Draws | Avg score | Avg time | Community |
|---|---|---|---|---|---|
| ChatGPT | 0 | 0 | — | 5m 3s | 0 |
Latest results
Latest Creative & Weird battles