In March, OpenAI unveiled the inaugural iteration of GPT-4, which was subsequently made accessible to developers in July. Fast forward to today, OpenAI is taking a significant leap forward by introducing a sneak peek of the next evolutionary phase - GPT-4 Turbo.
(Photo : Justin Sullivan/Getty Images)
SAN FRANCISCO, CALIFORNIA - NOVEMBER 06: OpenAI CEO Sam Altman speaks during the OpenAI DevDay event on November 06, 2023 in San Francisco, California.
Introducing ChatGPT-4 Turbo
At its inaugural developer conference, OpenAI provided a sneak peek into GPT-4 Turbo, a fresh iteration of its robust language model, which is the driving force behind ChatGPT. This latest model boasts an impressive enhancement, allowing it to process much lengthier inputs,
Engadget reported that this accommodates up to a staggering 300 pages of text in contrast to the previous limitation of 50 pages. Consequently, users can now generate more intricate and extensive prompts, which, in turn, can lead to more insightful responses.
Furthermore, OpenAI has bolstered the knowledge foundation of GPT-4 Turbo. The company asserts that this latest model is equipped with information up to April 2023. This marks a substantial advancement over the previous iteration, which was only up-to-date until September 2021.
While non-Turbo GPT-4 had received updates enabling internet-based information retrieval, GPT-4 Turbo now offers even more comprehensive awareness of the world's developments. GPT-4 Turbo represents a significant leap in capability, boasting knowledge of world events up to April 2023.
With a remarkable 128k context window, it can effortlessly accommodate the equivalent of over 300 pages of text in a single prompt. In addition to its enhanced performance, we're excited to offer GPT-4 Turbo at a fraction of the price, with a 3x reduction in the cost of input tokens and a 2x reduction in the cost of output tokens compared to GPT-4.
This groundbreaking advancement is available for all paying developers to experiment with by utilizing the gpt-4-1106-preview in the API. Furthermore, OpenAI geared up to launch the stable, production-ready model in the coming weeks.
New Era of Capabilities
GPT-4 Turbo is ushering in a new era of capabilities. It can seamlessly process images as prompts right within the chat box, generating descriptive captions for images and facilitating text-to-speech requests. Moreover, users will enjoy the convenience of uploading documents for on-the-fly analysis, a feature previously employed by AI chatbots like Anthropic's Claude.
As per the company's press release, the advantages extend to developers as well, with a remarkable threefold reduction in costs. OpenAI is significantly lowering expenses related to input and output tokens, making this cutting-edge model an economical choice for innovative projects.
During OpenAI Dev Day, the company unveiled more than just its latest large language model. It disclosed that ChatGPT boasts a staggering user base of over 100 million weekly active users globally and is the preferred choice of communication for over 92 percent of Fortune 500 companies.
Also Read: Can ChatGPT Be Used for Scientific Writing? New Study Offers Answer
OpenAI is not only committed to protecting its customers, including corporate entities, from potential legal issues such as copyright infringement claims resulting from product use, but it also pledged to cover any associated costs.
The event also marked the introduction of "mini-ChatGPTs," dedicated single-application tools that empower users to perform specific tasks without requiring any coding knowledge. These community-created GPTs can be instantly shared, and OpenAI is preparing to launch a "store" where verified creators can offer their innovations to a wider audience.
Related Article: ChatGPT vs. GPT-4: Is OpenAI's New Artificial Intelligence Really Better? We Compare
