OpenAI's cutting-edge GPT-4 API now accessible to all
On to some exciting news, OpenAI will now make the GPT-4 API available to paying customers signalling the retirement of older models in the coming months.
The ChatGPT API has witnessed an overwhelming response since its introduction in March, with millions of developers eagerly requesting access.
Existing paying customers can now start utilising the powerful GPT-4 API providing an 8K context for even more advanced AI capabilities as of July 6, 2023.
New developers will also gain access by the end of the month, which broadens the range of users to benefit from OpenAI's cutting-edge technology.
In response to the tremendous demand, OpenAI plans to raise rate limits based on compute availability. This will enable developers to harness the full potential of the GPT-4 API and explore its various applications.
The transition from the initial Completions API to the Chat Completions API has been a resounding success, with the latter accounting for a staggering 97% of OpenAI's API GPT usage. This shift reflects the company's commitment to improving the conversational experience and empowering users with enhanced interaction capabilities.
With the general availability of the GPT-4 API, older models utilising the Completions API are set to be deprecated, with the API labeled as 'legacy.' However, OpenAI reassures users that these legacy APIs will remain accessible, ensuring a smooth transition for developers.
To replace the older completion models, OpenAI will introduce ada-002, Babbage-002, curie-002, and DaVinci-002, offering even more advanced and efficient AI capabilities. Starting from January 4, 2024, these newer models will automatically replace their predecessors, ushering in a new era of AI technology.
OpenAI recognises that winding down the older models represents a significant change for developers who have relied on them. As a result, the company pledges to support affected users by covering the financial cost of re-embedding content with the new models. This commitment demonstrates OpenAI's dedication to its user community and ensures a smooth transition to the latest AI advancements.
The transition from a freeform text prompt to the Chat Completions API's structured prompt interface has been a key improvement. OpenAI believes that this approach leads to better results and a more enjoyable conversational experience. Additionally, the structured interface and multi-turn conversation capabilities of the Chat Completions API contribute to reducing the risk of prompt injection attacks. By adopting this approach, OpenAI enhances the security and privacy aspects of its AI systems, providing users with a safer and more reliable platform.
While OpenAI's AI tools have garnered significant popularity among companies seeking to integrate generative AI into their products, they have also faced legal challenges.
The company has been involved in multiple lawsuits and complaints, including a recent case in which it was named in a complaint filed in the Northern District of California US District Court concerning data usage in its models.
OpenAI is also facing a defamation lawsuit related to an erroneous accusation generated by one of its tools. Despite these legal challenges, OpenAI remains committed to improving its AI tools and ensuring they meet the highest ethical standards.