OpenAI Scraps o3 Model, Pushes for Unified GPT-5 in a Major AI Overhaul

OpenAI is canceling its o3 AI model and merging it into GPT-5 for a simpler, more powerful system. A big move to stay ahead in the AI race.

OpenAI has canceled its much-anticipated o3 reasoning model, opting to integrate its capabilities into the upcoming GPT-5. The decision, announced by CEO Sam Altman on X, marks a significant shift in OpenAI’s strategy as it seeks to simplify its product lineup and respond to growing competition in the AI industry. This OpenAI update reflects the company’s push to streamline development while enhancing overall performance.

Initially set for a February-March release, o3 was expected to be a breakthrough in reasoning AI, a model designed to fact-check itself and improve accuracy in complex tasks. As recently as last month, OpenAI’s Chief Product Officer Kevin Weil confirmed that o3 was on schedule. However, in an abrupt reversal, Altman stated that OpenAI would no longer release o3 as a standalone model, instead folding its technology into GPT-5.

We hate the model picker as much as you do and want to return to magic unified intelligence

 - Sam Altman 

Sam acknowledged user frustration over the complexity of OpenAI’s current AI offerings. This OpenAI update signals a major shift in strategy as the company consolidates its models to stay ahead of its rivals.

OPENAI ROADMAP UPDATE FOR GPT-4.5 and GPT-5:

We want to do a better job of sharing our intended roadmap, and a much better job simplifying our product offerings.

We want AI to “just work” for you; we realize how complicated our model and product offerings have gotten.

We hate…

— Sam Altman (@sama) February 12, 2025

GPT-5 to Launch With More Features, Free Access, and Tiered Intelligence

Before GPT-5 arrives, OpenAI will first release GPT-4.5, code-named “Orion,” in the coming weeks. Orion will be OpenAI’s last “non-chain-of-thought” model, meaning it won’t include self-fact-checking capabilities. Once GPT-5 launches, OpenAI will introduce tiered access: free-tier ChatGPT users will get unlimited chat at a “standard intelligence setting.” In contrast, Plus and Pro subscribers will access progressively more advanced versions.

GPT-5 is expected to integrate multiple AI capabilities, including voice mode, canvas, search, and deep research. These are the tools that OpenAI hopes will make its chatbot more intuitive and powerful. “A top goal for us is to unify o-series and GPT-series models by creating systems that can use all our tools, know when to think for a long time or not, and generally be useful for a wide range of tasks,” Altman stated.

With the OpenAI o3 model now merged into GPT-5, the company is aggressively staying competitive. This decision aligns with OpenAI’s broader mission to remain one of the best AI companies, ensuring its AI remains powerful and accessible. However, with rivals like DeepSeek pushing forward, the question remains: will a unified GPT-5 be enough to keep OpenAI ahead as one of the best AI companies?