Loading timeline…
20252020s
AI Achieves IMO Gold Medal
In July 2025, models from OpenAI and Google DeepMind both reached gold-medal standard at the International Mathematical Olympiad for the first time.
Why It Mattered
OpenAI's experimental reasoning model and Google DeepMind's Gemini Deep Think each scored 35 of 42 points, solving 5 of 6 problems under competition conditions — the first time AI crossed the IMO gold threshold, with DeepMind's result officially validated by IMO graders. Achieved with general-purpose language models rather than specialized formal-proof systems, the milestone demonstrated that natural-language mathematical reasoning had reached elite human level and foreshadowed AI's growing role in research mathematics.
Who Built It
OpenAI / Google DeepMind
Applications
- Mathematics
- Reasoning
- Scientific Research