OpenAI on Monday announced its most powerful language model yet - GPT 4 Turbo - at its first developer conference in San Francisco. The Sam Altman-led company said GPT-4 Turbo is more powerful and updated than the previous GPT-4 version, which was announced in March this year.
Speaking at the OpenAI Developer Conference, Altman announced that GPT-4 Turbo has knowledge of world events up to April 2023, a huge leap from the GPT-4 version that was trained on data up to September 2021. Earlier this year, however, OpenAI had introduced a feature that allowed ChatGPT to access the internet and get up-to-the-minute information.
The OpenAI CEO was quoted by Bloomberg as saying, “We are just as annoyed as all of you — probably more — that GPT-4’s knowledge about the world ended at 2021… We will try to never let it get that out of date again.”
Apart from being more up to date than the previous version, GPT-4 Turbo also features a larger 128K context window, which the company claims can fit over 300 pages of text in a single prompt. While previous versions of GPT-4 came with 8K and 32K context windows, GPT-4 Turbo even beats its competitors, the closest being Anthropic's Claude 2, which comes with a 100K context window.
A larger context window could mean that the large language model understands the question better and offers more thoughtful answers, reported The Verge.
In addition, OpenAI says GPT-4 Turbo will be cheaper for developers to use, with input tokens being three times cheaper at $0.01 and output tokens being two times cheaper at $0.03 compared to GPT-4.
In a blogpost about GPT-4 Turbo and other releases, OpenAI wrote, “We also optimized its performance so we are able to offer GPT-4 Turbo at a 3x cheaper price for input tokens and a 2x cheaper price for output tokens compared to GPT-4.”
Catch all the Business News, Market News, Breaking News Events and Latest News Updates on Live Mint. Download The Mint News App to get Daily Market Updates.