OpenAI has officially announced that its latest GPT-4.5 artificial intelligence model is now available to all ChatGPT Plus subscribers. According to OpenAI, it may take 1 to 3 days for the full rollout of the model, and users might experience slight server delays during the initial phase.

Performance comparison data indicates that GPT-4.5 demonstrates significant improvements over its predecessor, GPT-4o, across multiple areas. In professional inquiries, GPT-4.5 outperforms GPT-4o in 63.2% of cases. For everyday queries, this figure stands at 57%, while in creative tasks, GPT-4.5 surpasses GPT-4o by 56.8%.

The GPT-4.5 model was initially launched on February 28 as a "research preview" version. At the time, it was described as offering more natural interaction, an expanded knowledge base, enhanced understanding of user intent, and higher "emotional intelligence."

Prior to the current release, GPT-4.5 was made accessible to ChatGPT Pro users. Now, the model is progressively being rolled out to Plus and Team users, with plans to extend availability to Enterprise and Edu users in the future. Additionally, GPT-4.5 will also be deployed alongside other models on Microsoft’s Azure AI Foundry platform.

According to foreign media reports, OpenAI intends to launch the GPT-5 model by the end of May and is gradually phasing out the o3 model, incorporating its features into GPT-5.

It's worth noting that the cost of using GPT-4.5 is significantly higher than GPT-4o. Specifically, GPT-4.5 charges $75 per million tokens for input, compared to $2.5 for GPT-4o. For output, GPT-4.5 costs $150 per million tokens, whereas GPT-4o is priced at $10. This means the input cost of GPT-4.5 is 30 times higher than GPT-4o, and the output cost is 15 times greater.