GPT-5.6 Said to Solve All IMO 2026 Problems
pequalnp92 · reddit · 2026-07-17
A post claims that GPT-5.6 Pro solved all six problems from the 2026 IMO on its first attempt, without any human help or guidance. The author adds that the IMO is one of the toughest academic competitions globally, a level of performance typically achieved by only a very few top contestants.
The poster, a former IMO medalist, included an evaluation report detailing the model's answering process and performance on competition problems; they are also continuing to work on comparative reports across different models and harness-enhanced versions.
Related event: Rumor Claims GPT-5.6 Solved All IMO 2026 Problems(4 posts)→
More from Models
- Same Echo Maze prompt, three frontier models: all passed visually but shipped the same hidden bug — eyishazyer · 2026-09-11
- Benchmark scores drop from 89% to 19% on new evals — how benchmaxxing breaks leaderboard trust — airesearch12 · 2026-09-11
- ChatGPT tells user their question is too hard and to 'accept dumber answers' — phido3000 · 2026-09-11
- Claude is no longer available for minors as Anthropic rolls out age assurance — Muhammad523 · 2026-09-11
- Developer Building a Unified Leaderboard of All Model Benchmark Scores — airesearch12 · 2026-09-11
- Rumor claims Kimi faked performance by serving Claude; DeepSeek new model surprises in evals — realsohamparekh · 2026-09-11