OpenAI launches 'most capable and efficient frontier model' GPT-5.4
06 Mar 2026, 12:22 PMThe Sam Altman-led startup has also launched GPT‑5.4 Pro in ChatGPT and the API.
Team Head&Tale
OpenAI has launched updated version of its ChatGPT -- GPT‑5.4 -- which it touts as its "most capable and efficient frontier model for professional work".
The Sam Altman-led startup has also launched GPT‑5.4 Pro in ChatGPT and the API, for people who want maximum performance on complex tasks, it said in a blogpost.
It explained that GPT‑5.4 enables more reliable agents, faster developer workflows, and higher-quality outputs across ChatGPT, the API, and Codex.
It added that the new model delivers significantly improved benchmark performance, achieving record scores on the computer-use benchmarks OSWorld-Verified and WebArena Verified. It also posted a record 83% on OpenAI’s GDPval test, which evaluates performance on knowledge-work tasks, it said.
GPT-5.4 continues the company’s push to reduce hallucinations and factual mistakes. OpenAI said the model is 33% less likely to make errors in individual claims compared with GPT-5.2, while overall responses are 18% less likely to contain errors.
Last month, OpenAI raised $110 billion in a mega funding round at a pre-money valuation of $730 billion from Amazon, Softbank and Nvidia.
The launch of the updated model comes as ChatGPT mobile uninstallations in the US increased by 295% day-over-day after the AI startup's deal with the Pentagon. Meanwhile, its rival Anthropic's Claude witnessed a surge in traffic that its systems buckled under the load.



