Rumor: GPT-5.6 Aces the IMO

haider1 · x · 2026-07-17

Shared posts suggest unconfirmed rumors that GPT-5.6 might perfectly solve all 6 problems at IMO 2026.

The post notes that last year, OpenAI's experimental model and Google Gemini DeepThink both achieved gold-medal-level scores of 35/42, yet neither completely solved a single problem. This rumor underscores the fact that LLMs are advancing at an astonishing pace in mathematical competitions.

Related event: Rumor Claims GPT-5.6 Solved All IMO 2026 Problems(4 posts)→

Original post →

More from Models

Models channel →