OpenAI introduced several additions to its artificial intelligence models and services.
The firm unveiled new embedding models, application programming interface usage management tools and plans to reduce pricing for one of its models, according to a Thursday (Jan. 25) company blog post.
One of the features of OpenAI’s latest update is the introduction of two new embedding models: the text-embedding-3-small and the text-embedding-3-large, according to the post. These embeddings, which represent concepts within content, have undergone enhancements to improve performance and efficiency.
OpenAI has also updated its GPT models, particularly the GPT-3.5 Turbo model, the post said. This model, which has gained popularity among developers, will see a price reduction to facilitate increased use. The new GPT-3.5 Turbo model, gpt-3.5-turbo-0125, has improved accuracy in responding to requested formats and addresses a bug related to non-English language function calls.
In addition to these updates, OpenAI has updated a preview model of GPT-4 Turbo, gpt-4-0125-preview, per the post. The model is designed to provide more comprehensive task completion and reduce instances of incomplete or inadequate responses. OpenAI has also disclosed plans to launch GPT-4 Turbo with vision capabilities.
OpenAI has also made strides in moderation by releasing the text-moderation-007 model, which the company says is its most robust moderation model to date, according to the post. This model allows developers to identify potentially harmful text.
To enhance developers’ control and visibility over API use, OpenAI has introduced new tools, the post said. The tools allow developers to assign permissions to API keys, granting control over access to specific endpoints or read-only access for tracking purposes. Additionally, the usage dashboard and usage export function now provide metrics on an API key level, simplifying usage monitoring on a per-feature, -team, -product or -project basis.
The announcement of these updates comes a day after OpenAI was credited in part for Microsoft’s reaching a market valuation of $3 trillion. Microsoft’s partnership with OpenAI has helped Microsoft release AI-supported services to its customers at a time when investors have growing optimism about the potential of AI.
OpenAI’s own flagship product, ChatGPT Enterprise, has also gained traction in the corporate world, with 260 businesses signing up for the service within four months of its launch.
For all PYMNTS AI coverage, subscribe to the daily AI Newsletter.