Sam Altman reveals OpenAI's next global strategy for the first time

‍‍‍‍On February 12, OpenAI CEO Sam Altman said that OpenAI will launch its flagship product GPT-5 in the coming months and integrate its AI models into a more unified product.

01.GPT-4.5 and GPT-5 launch plan

Sam Altman published a series of content on the X platform, revealing the relevant schedule and sharing other details about the ChatGPT developer product roadmap.

Before the official release of GPT-5, a slightly less powerful large-scale language model, GPT-4.5, will be launched, which is expected to be launched in the next few weeks. This model is the successor to the GPT-4o model.

Although GPT-4o is not as good as OpenAI's large language model (LLM) with optimized reasoning ability in tasks such as generating code, it has a significant advantage in reasoning cost.

Altman pointed out in detail that GPT-4.5 will be OpenAI's last large language model without the Chain of Thought mechanism.

CoT is a machine learning technique that improves model performance by breaking down complex tasks into a series of simple steps. This method has been shown to significantly improve the output quality of LLM, especially when dealing with complex reasoning tasks.

Last year, sources told The Information that GPT-4.5's improvement in output quality would be far smaller than when its predecessor GPT-4 was released.

In addition, the upcoming new model reportedly did not perform “significantly better” than GPT-4 on programming tasks. This may be because OpenAI has put more effort into its inference-optimized LLMs, which are optimized for programming and mathematical tasks.

OpenAI’s latest inference-optimized LLM, o3, made its debut in December last year. It achieved a record high score of 25.2 percent on the "Frontier Math" benchmark test, which contains some particularly difficult math questions; the previous best score was less than 2 percent.

From Altman's X post, we learned that OpenAI has canceled its plan to release o3 as an independent product. Instead, the company plans to integrate the model into GPT-5, the next-generation model after GPT-4.5, which is expected to be launched in a few months.

Altman said: "One of our important goals is to unify the o series models and the GPT series models by creating systems that can use all tools. These systems can judge when they need to think for a long time and when they don’t, and are widely applicable to various tasks."

It is reported that OpenAI plans to provide GPT-5 in both free and paid versions of ChatGPT. Paid users will have access to a version of LLM with more advanced reasoning capabilities.

The new GPT-5 will also have built-in voice mode, canvas, search, and deep research functions, allowing chatbots to automatically adapt to user needs without users having to switch between different models and functions.

These advanced features are likely powered by OpenAI’s test-time computing technology. The technique improves the quality of the model's output by devoting more time and computing resources to processing the prompts. OpenAI has already applied this technology in o3.

02.OpenAI’s new strategy to cope with market competition

Last month, OpenAI also launched a tool called Deep Research, which is driven by o3 and aims to help users search for information on the Internet. It can generate lengthy reports based on data from hundreds of public sources.

Altman also revealed that free ChatGPT users will soon be able to run two Deep Research queries per month, while ChatGPT Plus subscribers will have the ability to run up to 10 queries per month.

Previously, OpenAI had maintained a leading position in the field of AI, but this position was recently challenged with the launch of DeepSeek's reasoning model R1.

The Chinese company reportedly developed the model at a very low cost and provides chat services for free, while its API access price is also much lower than OpenAI's model.

In addition to shaking the stock market, the rise of DeepSeek has also caused OpenAI and other AI companies to re-examine their business strategies, proving that AI models can be developed at a lower cost and with greater accessibility.

Perhaps Altman's announcement represents a shift in strategy.

Unlike its previous strategy of offering exclusive or early access to certain model features through its Pro plan for $200 per month, the new strategy seems to be more inclined to promote its products as widely as possible and beat competitors through market presence and user experience.

This adjustment may be in response to increasingly fierce market competition, especially competitors like DeepSeek that sell their products on low cost and high accessibility. By lowering the barrier to entry and expanding its user base, OpenAI may hope to consolidate its market position while attracting more users to rely on its ecosystem.