Recently, OpenAI made ChatGPT feel a bit more natural and less cringe with the GPT-5.3 Instant update. Now, the company is focusing on leveraging its AI’s raw intellectual power for professionals. OpenAI has officially launched GPT-5.4 Thinking and GPT-5.4 Pro, a duo of models designed to handle the heavy lifting that usually makes standard AI stumble.
It’s noteworthy that GPT 5.3 remains the default for your quick daily questions on ChatGPT. The new “Thinking” mode is where OpenAI is putting its most advanced reasoning capabilities.
You can control GPT-5.4 Thinking process
The most striking change for users is how ChatGPT now approaches a problem. Instead of jumping straight to a final response, GPT-5.4 Thinking outlines its plan upfront. According to OpenAI’s official announcement, this allows you to intervene in the middle of a task. If the AI starts planning a project and misses a crucial detail, you can steer it back on track without waiting for it to finish or starting the conversation over.
This “thinking out loud” capability makes the model significantly more reliable for long-horizon tasks. Now it should be even better for tasks such as planning complex travel itineraries or conducting deep web research. The model can stay focused on the final goal even during extended “reasoning stretches” by maintaining a stronger awareness of earlier steps in a conversation.
Built for professionals
OpenAI is clearly aiming for the office desk with this release. GPT-5.4 excels at “knowledge work”—tasks like spreadsheet modeling, document parsing, and presentation design. In internal tests, the model matched or exceeded human professionals in 83% of comparisons when producing specialized work across dozens of occupations.
For the data-heavy users, two features stand out. First is the built-in computer use capability of the model. It can interact directly with software, issuing mouse and keyboard commands based on screenshots to fix code or verify tasks. Then there is the massive context window that now supports up to 1 million tokens. This allows it to analyze entire codebases or massive document collections in a single go, a feature that directly competes with the latest offerings from Google and Anthropic.
Efficiency and accuracy
Beyond the new features, the model is simply more accurate. OpenAI reports that GPT-5.4 is 18% less likely to contain factual errors than previous versions. It is also more “token efficient,” meaning it solves complex problems using fewer resources. This translates to faster speeds and better performance on “tool-heavy” workloads—like when the AI needs to read emails, extract attachments, and record results in a spreadsheet all at once.
Availability
This update arrives at a curious time. Reports suggest that OpenAI has seen some user churn following a recent deal with the Pentagon. To counter this, OpenAI is leaning heavily into professional utility and performance. GPT-5.4 Thinking is rolling out now for Plus, Team, and Pro subscribers. Meanwhile, GPT-5.4 Pro targets Enterprise and API users who need maximum performance for high-stakes analytical work.
The post OpenAI’s New GPT-5.4 Thinking Lets You Take the Wheel Mid-Response appeared first on Android Headlines.
