Xiaohongshu's dots-note-3.0 Model Achieves Perfect Score at IMO 2026, First Chinese LLM Awarded Official Gold Medal

Xiaohongshu (Little Red Book) has made a stunning debut in the AI reasoning arena with its large language model dots-note-3.0 achieving a perfect score at IMO 2026. Officially evaluated by the International Mathematical Olympiad panel, the model solved all six problems correctly to earn a gold medal — the first time a Chinese LLM has received official IMO gold-level certification, and only the second model globally to achieve this feat after Google's Gemini.

This year's IMO gold medal threshold was 29 points, down from 35 last year, reflecting a significant increase in problem difficulty. dots-note-3.0's perfect score of 42 demonstrates not just gold-worthy performance but mastery at the maximum measurable level of the competition.

The model solved all problems using natural language end-to-end, reading the English problem statements and generating proofs directly without formal language translation. It employed an agentic approach combining natural language reasoning with Python code execution, and used recursive self-critique capabilities to validate its own arguments throughout the solving process.

Particularly impressive was the model's innovative approach to Problem 3, a combinatorial game theory challenge. While conventional solutions transform the problem into graph connectivity, dots-note-3.0 independently devised an induction-based proof that judges described as "elegant" and "structurally compact." Two-time CMO gold medalist Liu Hanzuo noted the model "produced a correct and beautiful proof," while Wang Qiantong praised the solution's "direct and essential" approach.

The IMO evaluation followed strict real-time procedures: the model team received each day's problems only after student competitors had completed their sessions, ensuring no prior data contamination. Final answer sheets were reviewed by IMO examiners from different countries using official standards.

For Xiaohongshu, primarily known as a content community platform, this represents a major technical coming-out party. The company plans to open-source the dots-note-3.0 model, offering the industry an opportunity to study its architecture and reasoning capabilities in depth. This move signals Xiaohongshu's emergence as a serious contender in the foundation model space.