[NEWS] OpenAI Announces Roadmap for GPT-4.5 and GPT-5

Sam Altman, CEO of U.S.-based OpenAI, announced a new roadmap for the next-generation AI models, GPT-4.5 and GPT-5, on the 13th. This update aims to advance AI model integration while enhancing the user experience.

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-4.5 is scheduled to be released within the next few weeks and will be the final "non-chain-of-thought model"provided by OpenAI. This model does not internally process complex reasoning steps but instead delivers only the final response. While retaining the strengths of GPT-4, it aims to improve performance and efficiency, serving as a transitional step toward the next-generation model.

On the other hand, GPT-5 is expected to be released within a few months and will integrate OpenAI's "o" series reasoning models, particularly o3, enabling it to handle a wider range of tasks. This integration is designed to provide users with a seamless experience, eliminating the need to choose between different AI models. Additionally, GPT-5 will be accessible in the free plan, while higher intelligence settings will be available in the paid plans, making advanced AI capabilities more widely available.

OpenAI acknowledges the increasing complexity of AI model choices in the current market, where a variety of models have emerged, making selection difficult for users. Moving forward, the company aims to create "magically integrated intelligence" that offers a more intuitive and user-friendly AI experience. The release of GPT-4.5 and GPT-5 represents a crucial step toward this goal, marking a significant advancement in AI technology while enhancing overall user convenience.

Since the beginning of this year, the Chinese startup DeepSeek has rapidly gained attention in the AI market. The company has developed low-cost, high-performance AI models, significantly impacting competitors such as OpenAI.

DeepSeek’s "DeepSeek-R1" model offers performance comparable to OpenAI’s GPT-4 while keeping development costs approximately one-tenth of OpenAI’s equivalent models. This exceptional cost-performance ratio has made it particularly attractive to small and medium-sized enterprises (SMEs) and startups.

Additionally, DeepSeek has adopted an open-source strategy, making its models freely available to promote the democratization of AI technology. This approach has enabled developers to leverage DeepSeek’s technology and foster an ecosystem where new AI applications can be built with ease.

As a result, the traditional notion that "high performance equals high cost" in the AI market is beginning to collapse, signaling another wave of transformation. The emergence of low-cost, high-performance AI models has forced major AI companies, including OpenAI, to reconsider their strategies. To stay competitive, OpenAI is accelerating the development of GPT-4.5 and GPT-5, likely releasing these next-generation models with higher intelligence and multimodal capabilities without significant delays.

As competition in the AI market intensifies, the rivalry between OpenAI and DeepSeek is expected to play a major role in shaping the next phase of technological innovation. With the growing adoption of open-source AI technology, the industry may soon witness an era where more advanced AI is made available at lower costs. As the AI market continues to evolve, OpenAI’s next strategic moves will be closely watched.

Reference: Announcement
Image: Quote from the announcement

Related articles

SoftBank Group in Talks for Additional Investment of Up to $25 billion in OpenAI

AI's Societal Advancement Seen Through Consecutive Nobel Prizes for AI Researchers