Perfect score achieved by an AI model in the International Math Olympiad (IMO) 2026?
Forecasting market: Perfect score achieved by an AI model in the International Math Olympiad (IMO) 2026?
Alpha Opportunity
Alpha Thesis
We believe the Manifold contract for AI achieving a perfect IMO score (42/42) in 2026 is overvalued at 64%, with our estimate at 30%. In 2025, the best AI systems scored 35/42 (gold medal level) — solving 5 of 6 problems. A perfect score requires solving ALL 6 problems, including Problem 6, which is traditionally the hardest and often goes unsolved by most gold medalists. Going from 5/6 to 6/6 is disproportionately harder than from 4/6 to 5/6.
📐Key Metrics
Key Findings
- 5/6 ≠ 6/6 — Solving the 6th (hardest) problem is qualitatively different from solving the easiest 5. Problem 6 often requires deep insight.
- Only 5 Humans Got 42/42 in 2025 — Out of hundreds of contestants. A perfect score is rare even for the best humans.
- IMO 2026 Is in Shanghai — July 15-16, 2026. Only 4 months away. AI systems need to be ready.
- AI Must Generate Formal Proofs — The IMO requires written proofs, not just answers. AI must produce rigorous, correct mathematical arguments.
- 30% Is Generous — Given that the gap between gold (35/42) and perfect (42/42) has historically taken humans years to close.
Full Research Report
Unlock the complete analysis including probability assessment, Bayesian calculations, resolution rigor analysis, and strategic positioning recommendations across 5+ dimensions.
Alpha Quality Factors
Criteria that determine how exploitable this mispricing is
Human Bias Detected
Cognitive biases creating this alpha opportunity
The crowd may lack specialized knowledge that narrows the true probability range.
Compare Markets
Searching Polymarket, Kalshi, Manifold & Metaculus…
Market Data
Position Sizing
Kelly Criterion (per $1,000 bankroll)