OpenAI is preparing to introduce new AI models — o3-mini and o3-mini-high, which will become part of the o3 series. These models are aimed at improving problem-solving in programming, science, and mathematics, providing faster response times compared to previous versions.
The key feature of the new models is their ability to handle complex queries with lower latency, making them suitable for both everyday and professional use.
OpenAI has been working on improving its reasoning models since the release of o1 in September 2024. The o3 series is a significant step forward, focusing on advanced reasoning benchmarks such as ARC-AGI and SWE-Bench Verified. While the full o3 model is targeted at applications with high computational demands, o3-mini is designed for users who need cost-effective yet powerful AI solutions.
The official launch of o3-mini is expected today, January 31, 2025, following Altman’s confirmation of a late January release. The model will be available in both free and paid ChatGPT tiers, with free-tier users facing stricter query limitations.
