What's happened
OpenAI will transition developers from GPT-4.5 to GPT-4o by July 14, 2025. GPT-4.5, while powerful, is costly and not the best option for all applications. The company is also preparing for the upcoming GPT-5, which promises enhanced capabilities.
What's behind the headline?
Key Insights
- Model Transition: OpenAI's decision to phase out GPT-4.5 in favor of GPT-4o reflects a strategic shift towards cost efficiency while maintaining performance.
- Cost Implications: GPT-4.5's operational costs are substantial, with pricing set at $75 per million input tokens, making it less viable for widespread use.
- Future Developments: The anticipated GPT-5 aims to integrate reasoning capabilities, indicating a trend towards more sophisticated AI interactions.
- Market Positioning: OpenAI's advancements come amid increasing competition from other AI firms, necessitating continuous innovation to retain market leadership.
- Resource Efficiency: The ability to retrain models with fewer resources, as highlighted by OpenAI's engineers, suggests a shift in operational dynamics that could lower barriers for future AI developments.
What the papers say
According to TechCrunch, OpenAI's CEO Sam Altman noted that GPT-4.5 will be available until July 14, 2025, after which developers must transition to GPT-4o, which offers similar performance at a lower cost. Business Insider UK reported that retraining GPT-4 has become significantly easier due to insights gained from GPT-4.5, with a much smaller team now capable of handling the task. This reflects a broader trend in AI development where efficiency and cost-effectiveness are becoming paramount. Additionally, Altman mentioned that GPT-5 is expected to be released in a few months, incorporating advanced features that could redefine user interactions with AI.
How we got here
OpenAI has been evolving its AI models, with GPT-4 launched in March 2023 and succeeded by GPT-4o. The introduction of GPT-4.5 in February 2025 marked a significant advancement, but its high operational costs have prompted a reevaluation of its API availability.
Go deeper
- What are the main features of GPT-4.5?
- How does GPT-4o compare to GPT-4?
- What can we expect from GPT-5?
Common question
-
What Are the New Memory Features in ChatGPT and How Will They Affect Users?
OpenAI has recently announced significant upgrades to ChatGPT's memory capabilities, allowing the AI to reference past conversations for a more personalized user experience. This change raises important questions about user interaction, privacy, and the overall effectiveness of the AI. Below, we explore common queries regarding these new features and their implications.
-
What are the key differences between GPT-4.5 and GPT-4o?
OpenAI is making significant changes in its AI offerings, transitioning developers from GPT-4.5 to GPT-4o. This shift raises questions about the differences between these models, the reasons behind the transition, and what the future holds for AI with the anticipated release of GPT-5. Here are some common questions and answers to help you understand these developments.
More on these topics
-
The United States of America, commonly known as the United States or America, is a country mostly located in central North America, between Canada and Mexico.
-
OpenAI is an artificial intelligence research laboratory consisting of the for-profit corporation OpenAI LP and its parent company, the non-profit OpenAI Inc.
-
Samuel H. Altman is an American entrepreneur, investor, programmer, and blogger. He is the CEO of OpenAI and the former president of Y Combinator.
-
GPT-4 is a multimodal large language model created by OpenAI, the fourth in the GPT series. It was released on March 14, 2023, and will be available via API and for ChatGPT Plus users.