OpenAI released the newest model

OpenAI has introduced GPT-4o Mini, a smaller and more affordable version of its GPT-4o large language model, now available through its cloud platform.

Announced on Thursday, GPT-4o Mini can handle various types of input, not just text. It has a context window of 128,000 tokens and was trained on data up to October 2023. The Mini version can generate up to 16,000 tokens of output.

Currently, GPT-4o Mini supports text and vision through its API, with additional input and output formats like audio expected to be added later.

OpenAI has also focused on making GPT-4o Mini safe to use. They have filtered out offensive data from its training materials and included the same safety features as GPT-4o. The Mini version also has mechanisms to prevent it from being tricked into doing things it's not supposed to do.

In our development process, we initially used GPT-3.5 Turbo for testing because it was cheaper. However, with the release of GPT-4o Mini, which is now even more cost-effective and smaller, we plan to switch to this new model for our projects.