OpenAI has introduced smaller AI models called embeddings to enhance applications' understanding of relationships between different content elements.
Alongside this, they have released updates, including the GPT-4 Turbo preview model, aimed at improving efficiency in tasks like code generation and reducing instances of 'laziness' in model performance.
OpenAI has also unveiled the gpt-3.5-turbo-0125 model as part of their offerings.
In response to customer needs, OpenAI has reduced the cost of GPT-3.5 Turbo API access for the third time in a year, halving input prices to $0.0005 per thousand tokens and lowering output prices by 25% to $0.0015 per thousand tokens, facilitating scalability for clients.
Additionally, they have introduced new functionalities for managing API keys and monitoring usage.