- GPT-4 will be phased out and replaced with the more advanced GPT-4o model in ChatGPT
- The new model outperforms GPT-4 in coding, science, math, and instruction-following tasks
- Users can access GPT-4o under the “ChatGPT-4o-latest” version, with upgrades for older API models also underway
OpenAI has announced that it will officially replace its GPT-4 model with the newer and more advanced GPT-4o by April 30, 2025.
The move marks a major step forward in AI performance, as GPT-4o is designed to deliver sharper capabilities, especially in technical tasks like coding, science, and mathematical problem-solving.
Backed by Microsoft, the AI firm shared the update on Thursday, April 10, noting that GPT-4o now consistently outperforms GPT-4 in various testing scenarios.
According to OpenAI, recent upgrades have significantly improved GPT-4o’s ability to follow instructions and maintain fluid, natural conversations.
These enhancements make the model more intuitive and capable for users across different fields.
With over 400 million people using ChatGPT weekly, the shift to GPT-4o means users will benefit from a more intelligent and reliable tool.
OpenAI emphasised that GPT-4o not only performs better with complex tasks but also generates cleaner front-end code, offers improved accuracy when identifying code issues, and provides solutions that compile and run more efficiently.
These improvements aim to streamline the coding experience for developers and tech enthusiasts.
"Recent upgrades have further improved GPT‑4o’s instruction following, problem-solving, and conversational flow, making it a natural successor to GPT-4.
“GPT-4o has further improved its capability to tackle complex technical and coding problems. It now generates cleaner, simpler frontend code, more accurately thinks through existing code to identify necessary changes, and consistently produces coding outputs that successfully compile and run, streamlining your coding workflows," OpenAI stated.
The company also highlighted that the new model has shown better results in creative tasks and now communicates with more clarity and conciseness.
Currently, GPT-4o is available in ChatGPT as the “ChatGPT-4o-latest” version and can also be accessed via the API.
OpenAI added that these enhancements would soon be extended to older models available through the API, although no specific timeline was provided.
OpenAI releases text-to-video model, Sora, for ChatGPT Plus and Pro users
Meanwhile, TheRadar earlier reported that OpenAI, the Microsoft-backed AI powerhouse unveiled its highly anticipated text-to-video generation tool, Sora, marking a significant milestone in generative AI technology.
Sora, now Sora Turbo, is a rebranded, faster, and more accessible tool designed for creators. It allows users to generate high-quality 1080p videos of up to 20 seconds in various aspect ratios, including widescreen, vertical, and square.