OpenAI has officially unveiled a simplified and more accessible version of its most powerful large language model, GPT-4o. The GPT-4o Mini neural network is designed to be more economically viable for developers while outperforming GPT-3.5.
Building applications using OpenAI language models can be costly. In the absence of sufficient funds, developers might abandon these models in favor of more affordable alternatives, such as Google’s Gemini 1.5 Flash or Anthropic’s Claude 3 Haiku. With the launch of GPT-4o Mini, competition in this segment is expected to intensify.
“I think GPT-4o Mini really fits with OpenAI’s mission to make AI more accessible to people. If we want AI to bring value to every corner of the world, every industry, every application, we must make AI much more accessible,” said an OpenAI representative.
Availability and Performance
Starting today, ChatGPT users on Free, Plus, and Team plans can use GPT-4o Mini instead of GPT-3.5 Turbo. Enterprise users will have access to this option next week. While the GPT-3.5 language model will no longer be available to ChatGPT users, it will still be accessible to developers via the appropriate API if they do not plan to migrate to GPT-4o Mini. The exact date when access to GPT-3.5 will be completely closed is still unknown, notes NIX Solutions.
The GPT-4o Mini language model supports text, images, audio, and video content. Although the algorithm is designed to solve simple problems, it can be useful for adding various narrowly focused functions to third-party applications. In the MMLU benchmark, OpenAI’s new language model scored 82%. For comparison, GPT-3.5 scored 70%, GPT-4o scored 88.7%, and Gemini Ultra scored 90%. Competing models, Claude 3 Haiku and Gemini 1.5 Flash, scored 75.2% and 78.9%, respectively.
We’ll keep you updated with further developments on GPT-4o Mini as more information becomes available.