OpenAI has announced the release of a new AI model that is at least 60% cheaper than its other chatbot models.
In a press release published Thursday on OpenAI's website, the company said it has released "GPT-4o mini," a new AI model that is part of the startup's effort to expand access to its popular ChatGPT bot. The "o" in GPT-4o mini stands for omni.
"The whole point of OpenAI is to build and distribute AI safely and make it broadly accessible," Olivier Godement, a product manager at OpenAI responsible for the new model, told WIRED. "Making intelligence available at a lower cost is one of the most efficient ways for us to do that."
The latest model is an offshoot of GPT-4o, which is currently OpenAI's fastest and most powerful model. Currently, the mini model supports text and vision in the API. The company plans to add all multimodal inputs and outputs to the models, including video and audio, as well as the ability to handle 40 different languages.
That being said, OpenAI noted that GPT-4o mini is only meant to be used for simple tasks.
How Much Will GPT-4o Mini Cost?
GPT-4o Mini is over 30 times cheaper than GPT-4o and 60% cheaper than GPT-3.5 Turbo, which is the company's most inexpensive existing model to date. The model has a context window of 128K tokens and supports up to 16K output tokers per request, according to the press release.
GPT-4o Mini is now available to free users of ChatGPT, along with Plus and Team subscribers. The new AI model will be available to Enterprise users beginning next week. GPT-4o Mini will replace GPT-3.5.
The release of GPT-4o Mini comes days ahead of the expected release of Meta's Llama 3 400B model---the largest version of its free offering, as first reported by The Information. Llama 3 is an open-source AI model that can be downloaded and used for free. However, there are certain limits on commercial usage.
The Llama 3 400B model is boasted to have 400 million parameters and is believed to be near-parity with OpenAI's GPT-4 model.
Join the Conversation