Contributed by the JetBrains community.

4 forecasters

For these benchmarks, what percentage of problems do you estimate the top-performing AI model or agent will be able to solve by December'25?

AI2 Reasoning Challenge
97.9
Graduate-Level Google-Proof Q&A
92.6
Toloka's µ-MATH
94.5
Toloka's U-MATH
91.5
Epoch's FrontierMath
47.5