(Photo by Jaap Arriens/NurPhoto via Getty Images)
OpenAI is launching GPT-4o Mini, a scaled-down and cheaper version of newest GPT-4o model.
"GPT-4o mini supports text and vision in the API, with support for text, image, video and audio inputs and outputs coming in the future," OpenAI says.
GPT-4o Mini replaces GPT-3.5 Turbo in ChatGPT and is available to those using the free and paid versions of OpenAI's chatbot. Enterprise users get it next week.
For developers, GPT-4o Mini will be a more affordable way to use OpenAI's tech. They'll pay $0.15 per million input tokens and $0.60 per million output tokens—60% cheaper than GPT-3.5 Turbo, OpenAI says. It has a context window of 128,000 tokens and is trained on data up to October 2023.
At I/O in May, Google also launched a smaller AI model “optimized for narrower or high-frequency tasks," dubbed Gemini 1.5 Flash.
At OpenAI, the company is aiming for a better multi-modal experience. "If you think about the way we as humans process the world and engage with the world, we see things, we hear things, we say things—the world is much bigger than text," OpenAI COO Brad Lightcap tells CNBC.
OpenAI launched GPT-4o earlier this year with promises of it being the most humanlike AI yet thanks to features like Voice Mode, which is currently delayed while OpenAI works out some kinks in the code. Look for it later this year.