OpenAI Unveils AI Models to Revolutionize Complex Problem Solving with 'Strawberry' Series

Microsoft-backed OpenAI has launched its 'Strawberry' series AI models, o1 and o1-mini, which excel in solving complex problems in science, coding, and math. These models incorporate 'chain-of-thought' reasoning, allowing them to break down problems into smaller steps. The models have shown significant improvements in various benchmarks.

OpenAI, backed by Microsoft, announced the launch of its 'Strawberry' series AI models, designed to tackle complex queries by spending more time processing answers. These models excel in science, coding, and math, far exceeding the capabilities of previous versions.

Known internally as 'Strawberry,' the models o1 and o1-mini will be integrated into ChatGPT and its API starting Thursday. Researcher Noam Brown shared that these models are the fruit of OpenAI's efforts to enhance general reasoning abilities.

The o1 model has achieved remarkable benchmarks, scoring 83% on the International Mathematics Olympiad qualifying exam and surpassing human PhD-level accuracy in science. The key to this success is the 'chain-of-thought' reasoning technique, which breaks down complex problems into manageable steps automatically.

(With inputs from agencies.)