Enterprise Generative AI SaaS 'HEROZ ASK' Now Supports OpenAI's Latest Model 'GPT-5.4' and Lightweight 'mini/nano' for High-Precision, High-Speed Operations

HEROZ, Inc. (Headquarters: Minato-ku, Tokyo; Representative Director and CEO: Takahiro Hayashi; hereinafter "HEROZ") has started offering OpenAI's latest AI model "GPT-5.4" and lightweight models "GPT-5.4 mini/nano" in its generative AI SaaS for enterprises, "HEROZ ASK." By selectively using models according to the application, we aim to achieve both quality improvement in highly specialized business support and response speed/operational efficiency in daily operations. ## Background of GPT-5.4 / GPT-5.4 mini/nano Support The utilization of generative AI in enterprises is rapidly expanding beyond document creation support such as proposals and reports, to include internal data search/summarization, inquiry response, and development operation support. On the other hand, in the field, different requirements are demanded for different applications, such as "complex examinations require high accuracy" and "daily routine tasks prioritize speed and cost efficiency." HEROZ will continue to quickly adapt to the latest models in its enterprise generative AI SaaS "HEROZ ASK," and provide customers with higher productivity and highly specialized business support. ## About GPT-5.4 GPT-5.4 is the latest large language model announced by OpenAI. It is expected to improve usability in practical business and provide stable answer quality in a wide range of applications such as text generation, reasoning, and coding. In "HEROZ ASK," it is assumed to be used for tasks involving multiple conditions or tasks that require organizing points of discussion. - Improved ability to handle complex requests (supporting condition sorting and clarification of points) - Boosting business-ready answer quality through stable output - Application to highly specialized business support ## About GPT-5.4 mini/nano GPT-5.4 mini/nano are positioned as lightweight and easy-to-handle model groups. They are expected to be utilized with an emphasis on response speed and operational efficiency in routine processing and high-turnover tasks. In "HEROZ ASK," they are assumed to be used for applications that reduce the burden of daily operations, such as short text summarization/classification and drafting primary responses. - Supporting high-speed drafting and organizing in daily operations - An option suitable for streamlining mass processing and repetitive tasks - Overall business optimization through application-specific operation