GPT-4.5, OpenAI’s most sophisticated AI language model to date, was released. Currently, ChatGPT Pro users and developers throughout the world can access the new model as a study peek.
The launch of ChatGPT’s most recent model follows the appearance of rival models like DeepSeek-R1, which have forced businesses like OpenAI to step up their innovation.
“This is the first model that feels like talking to a thoughtful person,” OpenAI CEO Sam Altman wrote in a post on the social networking platform X. Altman remarked, “I’ve sat back in my chair a few times and been amazed at getting really good advice from an AI.”
Altman further clarified that this model is “giant, expensive.” The CEO of OpenAI also clarified that benchmarks won’t be “crushed” by this technique. “There’s a magic to it that I haven’t experienced before, and it’s a different kind of intelligence,” he remarked.
What is new in GPT-4.5?
Early testing suggests that GPT-4.5 provides a more intuitive and natural user experience, according to OpenAI. According to a blog post by OpenAI, the model is especially good at jobs like writing help, programming, and real-world problem-solving because it has a larger knowledge base, better alignment with user intent, and more emotional intelligence. Additionally, according to OpenAI, GPT-4.5 should show less “hallucinations,” in which AI produces inaccurate or illogical information.
Nevertheless, GPT-4.5 is not categorized as a frontier AI model in spite of its improvements. In contrast to earlier models like o1 or o3-mini, OpenAI emphasizes that although the model improves user interface and computing efficiency, it does not bring completely new capabilities.
GPT-4.5 was created by combining new supervision techniques with more conventional approaches like reinforcement learning from human feedback and supervised fine-tuning. This method has produced a model that is more appropriate for group projects since it provides emotionally complex interactions in addition to lowering hallucinations.
GPT-4.5: Accessible
GPT-4.5 is first available to ChatGPT Pro users and developers. In the upcoming week, Plus and Team users will be added, followed by Enterprise and Education users. Furthermore, Microsoft’s Azure AI Foundry platform now incorporates the model, increasing its accessibility for business clients.