OpenAI introduces cheaper GPT-4o mini | Infinium-tech
OpenAI has just announced GPT-4o Mini — a new compact yet powerful AI model that aims to make the GPT-4 engine more affordable.
Its price is set at 15 cents per 1 million input tokens and 60 cents per 1 million output tokens, which is significantly cheaper than GPT-3.5 Turbo.
OpenAI revealed that the 4o Mini is much more powerful than other models such as Gemini Flash, Cloud Haiku, and GPT-3.5 Turbo. It outperforms all of them in benchmarks and produces results that are almost as good as GPT-4o.
The company explained in a press release that this smaller model focuses on textual intelligence and multimodal reasoning. GPT-4o outperforms other smaller models in mini reasoning tasks, excels at mathematical reasoning and coding, and also performs impressively for multimodal reasoning.
The engine has a reference window of 128k tokens and contains knowledge until October 2023. Image, video, and audio support is coming in the future.
The implementation of GPT-4o Mini begins today, replacing GPT 3.5 in Free, Plus, and Teams subscriptions for end users. Enterprise clients will be able to access the model starting next week.
Leave a Reply