OpenAI has introduced GPT-4o mini, a smaller version of its GPT-4o model. The company expects that the new product will significantly expand the range of applications created using AI, making artificial intelligence much more accessible.
Currently, the GPT-4o mini supports text and visual APIs, but the model's capabilities will be expanded over time. It has a context window for 128 thousand tokens and supports up to 16 thousand output tokens on request. The GPT-4o mini knowledge base is current as of October 2023.
The company also compared its product with other models. It turned out that in several key tests, the GPT-4o mini is ahead of Gemini Flash, Claude Haiku, and GPT-3.5 Turbo, but inferior to the GPT-4o.
ChatGPT users of the Free, Plus, and Team tariff plans can already access GPT-4o mini instead of GPT-3.5. Corporate customers will be able to try the product next week.
"We envision a future where models become seamlessly integrated in every app and on every website. GPT-4o mini is paving the way for developers to build and scale powerful AI applications more efficiently and affordably," noted at OpenAI.
In May 2024, OpenAI introduced its new GPT-4o artificial intelligence model, which significantly expands the speech processing capabilities of GPT-4.