Microsoft Corp.-backed (NASDAQ:MSFT) OpenAI has introduced the o3-mini, a new model in its reasoning series. It will be available on Friday in both ChatGPT and the API.

What Happened: The o3-mini model, previewed in December, promises to offer "exceptional capabilities" in science, math, and coding, while maintaining low costs and reduced latency.

It is the first small reasoning model from OpenAI to support developer features such as function calling and structured outputs, making it ready for production use.

According to OpenAI, the o3-mini will replace the o1-mini in the model picker, offering higher rate limits and lower latency. This makes it a compelling choice for tasks requiring coding, science, technology, engineering, and mathematics (STEM,) and logical problem-solving.

The model is rolling out to select developers in API usage tiers 3-5, with broader access for ChatGPT Plus, Team, and Pro users.

OpenAI emphasizes that o3-mini is optimized for STEM reasoning. Its medium reasoning effort matches the performance of the o1 model. It delivers faster responses, and evaluations show a 39% reduction in major errors on difficult questions compared to the o1-mini.

Why It Matters: The launch of the o3-mini model comes amid OpenAI's significant investments in infrastructure, such as the $100 billion Stargate data center in Texas.

Additionally, OpenAI CEO Sam Altman recently addressed the hype surrounding artificial intelligence, urging the public to temper expectations about the development of Artificial General Intelligence (AGI).

Altman's comments suggest that while OpenAI is making strides in AI capabilities, the journey toward AGI is still ongoing.

Furthermore, the introduction of the o3-mini follows the release of OpenAI's "Operator" AI agent, which was designed to autonomously perform web tasks. This agent, available to ChatGPT Pro subscribers, integrates advanced reasoning and vision capabilities, showcasing OpenAI's continuous innovation in AI technology.

Check out more of Benzinga's Consumer Tech coverage by following this link.

• OpenAI Debuts 'Operator' AI Agent That Performs Web Tasks, But Early Error Has Sam Altman Scrambling For A Quick Fix

Disclaimer: This content was partially produced with the help of Benzinga Neuro and was reviewed and published by Benzinga editors.

Photo courtesy: Shutterstock