OpenAI launched its latest model GPT-5.4 on Thursday. It is a new foundational model which OpenAI claims as its “most capable and efficient frontier model for professional work.” GPT-5.4 is also available as a reasoning model (GPT-5.4 Thinking) or optimized for high performance (GPT-5.4 Pro), in addition to the standard model.
The API version of the model will be available with context windows as large as one million tokens, by far the largest context window available from OpenAI.
OpenAI also emphasized improved token efficiency, saying GPT-5.4 was able to solve the same problems with significantly fewer tokens than its predecessor.
Read: OpenAI sees wave of senior exits following ChatGPT push
GPT-5.4 has improved benchmark results, including record scores in computer use benchmarks OSWorld-Verified and WebArena Verified. The new model also scored a record 83% on OpenAI’s GDPval test for knowledge work tasks. It also takes the lead at Mercor’s APEX-Agents benchmark, designed to test professional skills in law and finance, according to a statement from Mercor CEO Brendan Foody.
Foody said in a statement that “[GPT-5.4] excels at creating long-horizon deliverables such as slide decks, financial models, and legal analysis,” adding that it is “delivering top performance while running faster and at a lower cost than competitive frontier models.”
GPT-5.4 continues OpenAI’s efforts to limit hallucinations. The company said the new model was 33% less likely to make errors in individual claims when compared to GPT 5.2, and overall responses were 18% less likely to contain errors.
OpenAI has reworked how the API version of GPT-5.4 manages tool calling, introducing a new system called Tool Search as part of the launch. The new system allows models to look up tool definitions as needed, resulting in faster and cheaper requests in systems with many available tools.
OpenAI has also included a new safety evaluation to test its models’ chain-of-thought, the running commentary given by the models to show thought process through multi-step tasks. AI safety researchers have long worried that reasoning models could misrepresent their chain-of-thought, and testing shows it can happen under the right circumstances.
READ: OpenAI in talks for $60 billion investment from Amazon, Nvidia and Microsoft (January 29, 2026)
OpenAI’s evaluation reportedly shows that this is less likely to happen in the Thinking version of GPT-5.4, “suggesting that the model lacks the ability to hide its reasoning and that CoT monitoring remains an effective safety tool.”
This comes during a time of controversy for OpenAI with users increasingly deleting the ChatGPT mobile app after OpenAI signed a deal with the U.S. According to the statistics, after the news spread on Saturday, February 28, 2026, the app saw almost a 300% spike in uninstalls in the U.S. alone compared to the day before. The average rate of uninstallation is only 9%, which is a lot less compared to the single-day downfall.