The company claims these advanced models are capable of tackling complex tasks and solving more difficult problems in science, coding, and math compared to earlier versions.
The new models, OpenAI o1 and OpenAI o1-mini, are designed to think through problems more thoroughly before responding, mimicking the way a person would approach a challenging issue.
High-Level Performance Across Various Fields
OpenAI states that the o1 models perform at a high level, particularly on benchmark tasks in physics, chemistry, and biology. “These models function similarly to PhD students when faced with difficult challenges,” said the company. The models also excel in coding and mathematical problem-solving.
For instance, in a qualifying exam for the International Mathematics Olympiad (IMO), OpenAI’s GPT-4o managed to solve only 13% of the problems, while the new reasoning model achieved an impressive 83%.
The o1 model is expected to be useful in a variety of fields. OpenAI suggests it can assist healthcare researchers in annotating cell sequencing data, help physicists generate complex mathematical formulas required for quantum optics, and support developers in building and executing multi-step workflows.
Also Read: OpenAI Ventures for $150 Billion Valuation in Latest Funding Round
OpenAI o1-mini: A Cost-Effective Option
Along with the main o1 model, OpenAI has introduced the o1-mini—a faster and more affordable version, particularly designed for coding tasks. The company highlights that o1-mini is 80% cheaper than the o1-preview model, offering a cost-effective option for applications that require reasoning but not extensive world knowledge.
Availability of the OpenAI o1 Series
The o1 series models are now available for ChatGPT Plus and Team users, with the company rolling out access to end users. Both the o1-preview and o1-mini models can be manually selected from the model picker, with weekly message limits of 30 for o1-preview and 50 for o1-mini.
OpenAI is working on increasing these message limits and aims to enable ChatGPT to automatically choose the best model based on the given prompt.
Discover more from
Subscribe to get the latest posts sent to your email.
