Claude Fable, GPT-5.6 Sol, Kimi K3 all score 42/42 on IMO 2026

deedydas · x · 2026-07-21

The poster benchmarked Claude Fable, GPT-5.6 Sol, Kimi K3, and Axiom on the 2026 International Math Olympiad and found that all four reached a perfect 42/42. - **Claude Fable 5** solved the set in **1 attempt** and was the fastest at **2.5h** total. - **GPT-5.6 Sol** needed **one extra attempt** but was the cheapest in the comparison. - **Kimi K3** also solved everything, but required **more retries** and **much more token usage**. - **Axiom Math** proved the solutions in **Lean**. The image also shows that **P3 and P6** were the hardest problems by attempts and token count. The poster argues that the frontier has now moved beyond IMO math.

Related event: Multiple Frontier AI Models Achieve Perfect Scores in IMO 2026 Testing(5 posts)→

Original post →

More from Models

Models channel →