OpenAI has released a new AI model o3-mini with enhanced capabilities.

OpenAI has announced the release of a new version of its artificial intelligence model o3-mini, which is now available to ChatGPT users and via API. This new version has proven to be enhanced in the fields of science, mathematics, and programming, while maintaining an affordable usage cost and reduced latency compared to its predecessor o1-mini.

OpenAI o3-mini has three levels of reasoning - low, medium, and high, allowing users to customize its operation for specific tasks. According to the company, the new model is characterized by a response speed that is 24% faster than o1-mini - on average 7.7 seconds compared to 10.16 seconds.

ChatGPT Plus and Team users have received increased message limits from 50 to 150 per day, while Pro account holders now have unlimited access to both the standard o3-mini version and the upgraded o3-mini-high version. It is worth noting that even free ChatGPT users can now utilize the model with reasoning capabilities.

The trials of o3-mini showed higher accuracy compared to o1-mini. Experts preferred the answers from the new model in 56% of cases and noted a 39% decrease in severe errors when solving complex tasks.

Risks and opportunities of artificial intelligence

It has been noted that artificial intelligence may pose a threat to the financial sector due to an increased potential for fraud and misinformation in financial markets.

There are also concerns about the use of artificial intelligence in the military sphere, as during simulations of military conflicts, the model showed a tendency to use nuclear weapons.

However, powerful artificial intelligence can also accelerate scientific progress and allow for the achievement of the next 50-100 years of biological development in just 5-10 years.