OpenAI will soon phase out GPT-4 from ChatGPT

OpenAI has announced that it will soon retire GPT-4, the AI model launched over two years ago, from ChatGPT. According to a changelog posted on Thursday, GPT-4 will be "fully replaced" by GPT-4o effective April 30. However, GPT-4 will remain available via OpenAI’s API.

Why GPT-4o?

OpenAI highlighted that GPT-4o consistently outperforms GPT-4 in evaluations across writing, coding, STEM, and other domains. Recent upgrades have further enhanced GPT-4o’s capabilities in instruction following, problem-solving, and conversational flow, making it a natural successor to GPT-4.

GPT-4’s Legacy

GPT-4 was initially rolled out in March 2023 for ChatGPT and Microsoft’s Copilot chatbot. It was notable for its multimodal capabilities, enabling it to process both text and images—a first for a widely deployed OpenAI model. OpenAI CEO Sam Altman revealed that training GPT-4, reportedly a massive model, cost over $100 million. It was later succeeded by GPT-4 Turbo in November 2023, a faster and more cost-effective version.

Legal and Future Implications

GPT-4 is central to ongoing copyright disputes between OpenAI and publishers, including The New York Times. Publishers allege OpenAI trained GPT-4 on their data without consent, while OpenAI argues fair use protects it from liability.

What’s Next?

OpenAI is reportedly preparing new models, including a family of GPT-4.1 variants (GPT-4.1-mini, GPT-4.1-nano, and GPT-4.1) and the o3 reasoning model, announced in December. A new model, o4-mini, is also in the works, according to reverse engineer Tibor Blaho.

About the Author

Kyle Wiggers is TechCrunch’s AI Editor. His work has appeared in VentureBeat, Digital Trends, and gadget blogs like Android Police and XDA-Developers. View Bio.