OpenAI launches cheaper and smarter mini GPT-4o

 This is part of the company's efforts to be at the forefront of multi-modality, offering a variety of AI-generated media within one engine such as text, images, audio and video.

OpenAI launches cheaper and smarter mini GPT-4o Picture 1

GPT-4o mini is a fork of GPT-4o, the fastest and most powerful model released by OpenAI last May. GPT-4o is described as having improved audio, video and text generation, with the ability to handle 50 different languages, while improving speed and quality.

George Cameron, co-founder at Artificial Analysis, said the GPT-4o mini is fast, with an average output rate of 202 tokens per second.

The GPT-4o mini will replace the GPT-3.5 Turbo as the smallest model OpenAI offers and begins rolling out to free ChatGPT users, ChatGPT Plus and Team subscribers, available on ChatGPT Enterprise next week.

GPT-4o mini is significantly cheaper than previous AI models and 60% cheaper than GPT-3.5 Turbo, at 0.15 USD per million input tokens and 0.60 USD per million tokens output message.

Marvin Fry

Update 19 July 2024