On Wednesday, OpenAI unveiled an upgraded iteration of GPT-4 Turbo, which is now available through an Application Programming Interface (API) and for ChatGPT Plus subscribers, as reported by multiple media outlets. This new version of GPT-4 Turbo showcases OpenAI’s latest advancements in generative AI, having been trained with data up to December 2023.
According to The Indian Express, the improved model now has enhanced data processing capabilities. It can analyze text in JSON format and perform function calls to create JSON code snippets. These snippets allow developers to automate various tasks within their connected applications, including sending emails, posting online, or making purchases.
As per the national daily, developers now have the ability to utilize the expanded context window of up to 128,000 tokens in the new API. This allows them to deploy GPT-4 Turbo for tasks such as image analysis, including food recognition across diverse global cuisines. Moreover, the model exhibits improved skills in generating computer code. Developers who have access to the GPT-4 Turbo API can effortlessly integrate this latest model into their applications and services.
The enhanced GPT-4 Turbo can now create complete and functional websites by generating authentic code and interpreting image content and context to produce meaningful results.
The GPT-4 Turbo version with vision capabilities via JSON mode is now available to ChatGPT Plus subscribers for $20 per month. This edition can also be obtained through various editions of ChatGPT on the ChatGPT store. Users have the option to upgrade to ChatGPT-4 Turbo using Microsoft’s Copilot, selecting either the creative or precise mode.
Back in January of this year, OpenAI, led by Sam Altman, rolled out a series of updates to its AI models and pricing to tackle increasing competition. This included a reduction in pricing for GPT-3.5 Turbo and enhancements to the performance of GPT-4 Turbo in response to user feedback about the model’s perceived “laziness.”
As per IE’s report, while the industry awaits the launch of GPT-5 with anticipation, other models have made notable progress in matching or even surpassing the abilities of their predecessors. Anthropic’s Claude 3 Opus, which debuted in March, has taken the lead in the closed-source models leaderboard, surpassing GPT-4. Following closely is Google’s Gemini Pro.
Source: BS