Claude Opus 5 reportedly scores a perfect 42/42 on the 2026 IMO

exordin26 · reddit · 2026-07-25

• The image shows a benchmark-style report claiming Claude Opus 5 achieved a perfect 42/42 on the 2026 IMO.
• According to the report, the model solved all six problems and produced four independent solutions per problem.
• The evaluation was done with three frontier-model judges; a solution counted only when all three agreed it was correct.
• The report says the score corresponds to gold-medal performance, well above the 2026 gold cutoff of 29/42.
• The screenshot also notes the run used no agent harness or tools, a 256,000-token output limit, and adaptive thinking at max effort; one problem required resampling after hitting the output limit.

Original post →

More from Models

• Opus 5 System Card Reveals Algebra Conversion Tactic for ARC-AGI Puzzles — herbiebradley · 2026-07-25
• GPT-5.6 beat Claude Opus 5 on landing-page polish in a frontend test — Arindam_1729 · 2026-07-25
• Google is called out for weak Gemma models and its stance on open-source LLMs — BoogerheadCult · 2026-07-25
• Musk says Grok 4.6 is due in 2 weeks and Grok 4.7 in 4 weeks — elonmusk · 2026-07-25
• Together says Kimi K3 matches near-flagship coding at about 35% of Claude Fable 5’s price — togethercompute · 2026-07-25
• Musk says Grok 4.5 and Claude Opus 5 are the only models on the Pareto frontier — elonmusk · 2026-07-25

Models channel →