According to news on July 10, OpenAI yesterday announced the full opening of GPT-3.5 Turbo, DALL-E and Whisper APIs to assist developers in improving model processing efficiency. In addition, OpenAI also stated that it is working on Develop follow-up features for GPT-4 and GPT-3.5 Turbo, which are planned to be launched in the second half of this year.
OpenAI revealed that all current AI models called by the API have been upgraded to GPT-4 by default, and existing users can use it without switching.
Note: Whisper API is a speech-to-text AI model that can recognize the user's voice, video and other media and convert it into text.
▲ Picture source OpenAI official website
OpenAI stated that it is constantly improving the Chat Completions API, with the main goal of improving its computing efficiency. They plan to discontinue the obsolete model of the Completions API in January 2024, 6 months later.
Note: Completions API is a natural language processing API that can be used for various text generation tasks, such as generating summaries, translating languages, generating articles, self-service Q&A, etc.
OpenAI claims that developers can still call the Completions API, but starting today, OpenAI will mark "old versions of Completions" as "old APIs" in developer files. OpenAI will focus resources on the Chat Completions API in the future and will no longer expose models using the Completions API.
IT House learned after inquiries that OpenAI stated that it is expected that by January 4, 2024, the old model will be terminated, and the Edits API will also be deactivated. Users of the Edits API and related models must be deactivated in early January next year. From the old model (text-davinci-edit-001 or code-davinci-edit-001) to GPT-3.5 Turbo, OpenAI also lists a comparison table of the old and new models.
▲ Picture source OpenAI
The above is the detailed content of OpenAI has fully opened GPT-3.5 Turbo, DALL-E and Whisper APIs. For more information, please follow other related articles on the PHP Chinese website!