In a groundbreaking announcement, OpenAI has introduced a host of thrilling updates to its API models, offering developers enhanced capabilities, increased flexibility, and more affordable pricing options.
Hot off the heels of the successful launch of gpt-3.5-turbo and gpt-4 earlier this year, OpenAI has witnessed an astonishing array of applications built by developers utilizing these cutting-edge models. Building on this momentum, OpenAI is now delighted to share some exciting news:
First and foremost, developers can now take advantage of the new function calling capability in the Chat Completions API, courtesy of OpenAI. This revolutionary addition allows seamless integration of external tools and APIs with models like gpt-4-0613 and gpt-3.5-turbo-0613. With this advanced feature, developers can intelligently generate JSON objects containing arguments for function calls, forging a stronger connection between GPT's capabilities and external tools.
But that's not all! OpenAI has also unveiled updated and more controllable versions of gpt-4 and gpt-3.5-turbo. These models have been meticulously fine-tuned to offer developers increased steerability, allowing them to guide the model's responses more effectively. This empowers developers to tailor the output to meet their specific requirements, enhancing the overall quality and usefulness of the models.
In response to the demand for handling larger text inputs, OpenAI has gone the extra mile by introducing a new 16k context version of gpt-3.5-turbo. With an extended context length, approximately four times that of the standard version, developers can now process and generate responses for around 20 pages of text in a single request. This exciting enhancement unlocks a world of possibilities, enabling more efficient handling of large-scale text-based tasks.
As part of OpenAI's commitment to making AI more accessible, significant price reductions have been implemented across various offerings. The highly popular embeddings model, text-embedding-ada-002, now comes with a staggering 75% reduction in cost, making it more affordable than ever at just $0.0001 per 1K tokens. Furthermore, OpenAI has reduced the cost of gpt-3.5-turbo's input tokens by 25%. Developers can now leverage this powerful chat model for a mere $0.0015 per 1K input tokens and $0.002 per 1K output tokens, providing exceptional value for their money.
In line with OpenAI's commitment to continuous improvement, the company is initiating the upgrade and deprecation process for the initial versions of gpt-4 and gpt-3.5-turbo, initially announced in March. Starting from June 27th, applications using the stable model names (gpt-3.5-turbo, gpt-4, and gpt-4-32k) will be automatically upgraded to the new models mentioned above, ensuring a seamless transition for developers. OpenAI understands that some developers may require additional time to transition, and as such, the older models (gpt-3.5-turbo-0301, gpt-4-0314, and gpt-4-32k-0314) will remain accessible until September 13th. However, after this date, requests specifying these model names will no longer be supported.
OpenAI remains committed to data privacy and security. As per their guarantee introduced on March 1, customers retain ownership of all outputs generated from their requests, and their API data will not be used for training purposes.
OpenAI is thrilled to invite developers from the waitlist to explore the capabilities of the latest gpt-4 model in the coming weeks. The aim is to eliminate the waitlist entirely with this model, providing equal opportunities to all developers eager to embark on their AI-powered journey with GPT-4.
With these remarkable updates and OpenAI's unwavering dedication to improving the platform based on developer feedback, the possibilities for innovation in AI-powered applications are endless. OpenAI eagerly awaits witnessing the astounding creations and achievements of developers harnessing the power of these cutting-edge models and features.
Stay tuned for further updates and advancements as OpenAI continues to push the boundaries of AI, equipping developers with the tools they need to revolutionize the world through the power of language.
Post a Comment
Post a Comment