A new model, GPT-5.4, has been launched by OpenAI after weeks of collaboration with research, product teams, and API customers.
It focuses on better performance in long-running tasks, improved control over style/behavior, and more reliable execution in complex workflows.
Early results show strong improvements in areas like agents, coding, computer use, and economically valuable tasks.
A new Prompting Guide was also released to help developers get more value from the model.
Link: https://lnkd.in/eSWYsxYc