According to multiple media reports, OpenAI released an improved version of GPT-4 Turbo on Wednesday. It is available to ChatGPT Plus users as well as through an Application Programming Interface (API). The company’s most recent generative AI model, GPT-4 Turbo, was trained using data through December 2023.
With its enhanced data processing capabilities, the upgraded model can now analyze text in JSON format and generate JSON code snippets by invoking functions. According to The Indian Express, these snippets enable developers to automate processes like emailing, posting online, and making purchases within their connected applications.
Developers can use the larger context window, which can hold up to 128,000 tokens, to apply GPT-4 Turbo for picture analysis, which includes the recognition of food from different international cuisines. Furthermore, the model shows improved ability to write computer code. According to the national-daily, developers that have access to the GPT-4 Turbo API can easily incorporate the most recent model into their services and apps.
With the latest update, the GPT-4 Turbo can now create fully working websites by generating genuine code and understanding image content and context to produce useful output.
ChatGPT Plus users can now use the version of GPT-4 Turbo that has vision capabilities via JSON mode for $20 a month. Popular ChatGPT editions are also selling this version through the ChatGPT store. Through Microsoft’s Copilot, users can update to ChatGPT-4 Turbo and select between the creative and precise modes.
Under Sam Altman’s guidance, OpenAI made a number of changes to its AI models and price structure in January of this year in response to growing competition. GPT-3.5 Turbo’s price was lowered, while GPT-4 Turbo’s performance was increased in response to user concerns about what they saw as the model’s “laziness.”
Even if the industry can’t wait for GPT-5 to launch, other models have made great progress in matching or even surpassing the capabilities of their predecessors. According to IE, Anthropic’s Claude 3 Opus, which was published in March, has overtaken GPT-4 on the scoreboard for closed-source models. Google’s Gemini Pro is in tight second place.
