OpenAI has announced a major update to its language model APIs, including GPT-4 and GPT-3.5-turbo, with new function-calling features, lower costs, and an option for the gpt-3.5-turbo model to support a 16,000-token context window. The context window in language models is like short-term memory that saves the contents of the prompt input or conversation in chatbots. OpenAI has released the gpt-3.5-turbo-16k model, which is four times larger than the standard 4,000 version, and it can process up to 20 pages of text in one request.
The company listed other major updates in its GPT APIs, including the ability to call functions, better steerability, reduced pricing for the ada embeddings model by 75%, and a 25% reduction in input tokens for the gpt-3.5-turbo model. With function calling, developers can make apps that call external tools, turn natural language into API calls or query a database. Steerability refers to the ability to get the language model to act as desired, and the new gpt-3.5-turbo-0613 model offers more reliable steerability.
Lower costs for the models will allow developers to use them more frequently. The price of input tokens for the gpt-3.5-turbo model has gone down by 25%, making it more affordable at $0.0015 per 1,000 input tokens and $0.002 per 1,000 output tokens. Furthermore, the text-embedding-ada-002 model’s price has been reduced by 75%. In addition to these updates, OpenAI has made it clear that it will no longer be making older versions of these models such as the gpt-3.5-turbo-0301 and gpt-4-0314. Developers can still use these models until September 13th, after which they will be phased out.
Frequently Asked Questions (FAQs) Related to the Above News
What updates has OpenAI announced to its language model APIs?
OpenAI has announced updates to its language model APIs including GPT-4 and GPT-3.5-turbo, with new function-calling features, lower costs, and an option for the gpt-3.5-turbo model to support a 16,000-token context window.
What is a context window in language models?
The context window in language models is like short-term memory that saves the contents of the prompt input or conversation in chatbots.
What is the gpt-3.5-turbo-16k model?
The gpt-3.5-turbo-16k model is four times larger than the standard 4,000 version, and it can process up to 20 pages of text in one request.
What is the function-calling feature added to the OpenAI APIs?
With function calling, developers can make apps that call external tools, turn natural language into API calls or query a database.
What is steerability and how has it been updated in the GPT APIs?
Steerability refers to the ability to get the language model to act as desired, and the new gpt-3.5-turbo-0613 model offers more reliable steerability.
Has OpenAI reduced the pricing of its models?
Yes, OpenAI has reduced the pricing of its models, including a 25% reduction in input tokens for the gpt-3.5-turbo model and a 75% reduction in the price of the text-embedding-ada-002 model.
Can developers still use older versions of the OpenAI models?
Developers can still use older versions of the OpenAI models such as the gpt-3.5-turbo-0301 and gpt-4-0314 until September 13th, after which they will be phased out.
Please note that the FAQs provided on this page are based on the news article published. While we strive to provide accurate and up-to-date information, it is always recommended to consult relevant authorities or professionals before making any decisions or taking action based on the FAQs or the news article.
