As DeepSeek builds up heat, OpenAI releases a cost-efficient reasoning AI model; here's what it can do

Just as China’s DeepSeek started to gain attention for its R1 model, OpenAI wasted no time to rolling out an affordable and efficient reasoning model, "o3-mini", according to a report.

The model was previewed in December but released on Friday, the same week when DeepSeek's open-source R1 model disrupted big tech, reported Business Insider.

Affordable reasoning

The new o3-mini is designed to handle complex queries in science, math, and coding, and allows developers to adjust the model’s reasoning complexity to suit different levels of difficulty. The model claims to be an affordable alternative to the higher-priced AI solutions currently in the market.

Available from today

It is available for users who have ChatGPT Plus, Team, and Pro. Free plan users can try it out with the "Reason" option, as per the report.

DeepSeek's investigation

This launch comes at a time when OpenAI is investigating whether DeepSeek may have used distillation techniques to replicate US-based AI models. DeepSeek’s open-source R1 has raised eyebrows due to its claimed cost-effectiveness and efficiency compared to established models like OpenAI’s own offerings.

SoftBank to invest

While many tech giants, including Meta and Microsoft, continue to double down on massive investments in AI, OpenAI is reportedly in discussions for a $25 billion investment from SoftBank, which could bring its valuation to $300 billion, reported Business Insider.

FAQs

What makes OpenAI’s o3-mini model different?
o3-mini is OpenAI's most cost-efficient reasoning model, designed to handle complex tasks in areas like science, math, and coding. It allows developers to adjust the complexity of the reasoning.

How can I access o3-mini?
o3-mini is available to ChatGPT Plus, Team, and Pro users. If you're on the free plan, you can try it out with the "Reason" option to explore its capabilities.