OpenAI launches affordable GPT-4o mini

OpenAI, the maker of ChatGPT, announced on Thursday the release of GPT-4o tiny, a cost-effective compact AI model targeted at making its technology more economical and energy-efficient, allowing the firm to reach a larger customer base.

Microsoft-backed (MSFT.O). OpenAI, the market leader in AI software, has been pushing to make it cheaper and faster for developers to build apps based on its model, even as deep-pocketed rivals like Meta (META.O) and Google (GOOGL.O) compete for a larger portion of the market.

The GPT-4o small costs 15 cents per million input tokens and 60 cents per million output tokens, making it more than 60% less expensive than the GPT-3.5 Turbo, according to OpenAI.

According to OpenAI, it now outperforms the GPT-4 model in chat preferences and achieved an 82% score in Massive Multitask Language Understanding (MMLU).

MMLU is a textual intelligence and reasoning benchmark designed to assess the capabilities of language models. A higher MMLU score indicates that it can grasp and use language more effectively across multiple domains, which improves real-world performance.

According to OpenAI, the GPT-4o small model scored 77.9%, while Google’s Gemini Flash and Anthropic’s Claude Haiku scored 73.8% each.

Smaller language models require less processing power to run, making them a more cost-effective solution for businesses with limited resources wishing to incorporate generative AI into their operations.

With the micro model now supporting text and vision in the application programming interface, OpenAI announced that text, picture, video, and audio inputs and outputs would be added in the future.

ChatGPT’s Free, Plus, and Team users will be able to access GPT-4o mini, which has knowledge until October 2023, beginning Thursday, replacing GPT-3.5 Turbo, with enterprise users obtaining access next week, according to OpenAI.