Math on Best of AI

AlphaGeometry

Thu, 16 Jul 2026 00:00:00 +0000

AlphaGeometry is a neuro-symbolic system from Google DeepMind, published in Nature in January 2024, built to solve geometry problems at the level of International Mathematical Olympiad competitors. It pairs a language model trained on synthetically generated geometry proofs with a traditional symbolic deduction engine: the neural half proposes useful auxiliary constructions, such as new points or lines, and the symbolic half runs formal logical deduction to complete the proof. On a benchmark of 30 IMO geometry problems from 2000 to 2022, it solved 25, close to the average gold medalist’s score and far ahead of prior automated theorem provers.

DeepSeek-Math

Thu, 16 Jul 2026 00:00:00 +0000

DeepSeekMath is a 7-billion-parameter model DeepSeek built specifically for mathematical reasoning, released in February 2024 and initialized from DeepSeek-Coder-Base. It was trained on 120 billion math-related tokens pulled from web data, alongside natural language and code, and scored 51.7% on the competition-level MATH benchmark without relying on external calculators or tools, a result that at the time approached the performance of much larger closed models like Gemini-Ultra and GPT-4. DeepSeek released it in base, instruction-tuned, and reinforcement-learning-tuned variants, with the RL version trained using Group Relative Policy Optimization, an algorithm the company introduced in this paper and later reused for DeepSeek-R1. The model has a 4,096-token context window and is aimed at researchers and developers working on math tutoring, theorem proving, and quantitative reasoning tasks rather than general-purpose chat.

LLeMA-7B

Thu, 16 Jul 2026 00:00:00 +0000

Mathstral

Thu, 16 Jul 2026 00:00:00 +0000

o3

Thu, 16 Jul 2026 00:00:00 +0000

Orca-Math

Thu, 16 Jul 2026 00:00:00 +0000