The tests used to measure and compare AI models, from math and coding to agents and real-world tasks. They're how labs show progress and how users choose a model. As models master each test, new and harder ones keep appearing, and how fair they are is often debated.
1,158 stories so far in 2026
Stories per month in 2026| Month | Stories |
|---|
| January | 2 |
|---|
| February | 0 |
|---|
| March | 344 |
|---|
| April | 185 |
|---|
| May | 26 |
|---|
| June | 169 |
|---|
| July | 62 |
|---|
| August | 67 |
|---|
| September | 95 |
|---|
| October | 136 |
|---|
| November | 72 |
|---|