Last week, Sam Altman, CEO of OpenAI, announced the company’s decision to forgo the release of the o3 artificial intelligence (AI) model. In January, OpenAI had introduced the o3-mini model to all ChatGPT users, and the more advanced o3 model was anticipated for launch this month. However, while outlining the company’s AI roadmap, Altman indicated that the focus would instead shift to a consolidated GPT-5 model that will encompass the functionalities of o3.
OpenAI to Launch GPT-4.5 Before GPT-5
In a recent post on X, formerly known as Twitter, Altman provided insights into the plans for GPT-4.5 and GPT-5. The upcoming two models aim to streamline OpenAI’s product offerings, lessening the complications users face when choosing models for various tasks.
Altman further explained that starting with GPT-5, the organization intends to deliver “unified intelligence.” This integrated model will leverage all available OpenAI tools, allowing users to engage without having to discern which model is best suited for specific tasks.
To implement this vision, the capabilities of o3 will be incorporated into the GPT-5 AI model, indicating that a standalone release of the o3 model will not take place. OpenAI appears to be moving toward phasing out independent releases of the o-series AI models.
Alongside these changes, the firm is revising its tiered service strategy with the introduction of GPT-5. Altman noted that this model will be accessible to all ChatGPT users upon launch, with free-tier users enjoying unlimited chat access at a standard intelligence level, contingent on the company’s abuse prevention measures.
Subscribers of ChatGPT Pro will benefit from operating GPT-5 at a superior intelligence level relative to the free version, with Pro users receiving the highest intelligence capabilities. While the term “intelligence” has not been precisely defined, the company is likely maintaining ambiguity around the specific capabilities for each AI model. Nevertheless, Altman did disclose that GPT-5 will feature Advanced Voice, Canvas, Search, Deep Research, and additional functions.
Before these advancements are rolled out, OpenAI plans to launch the GPT-4.5 model, which is expected to be the last of its kind that does not utilize chain-of-thought reasoning. Although Altman did not specify a release date, he mentioned that the model could debut in the coming weeks or months. Additionally, he revealed that GPT-4.5 is internally referred to as Orion, a name previously speculated to be associated with GPT-5.
