Japanese firms launch new generative AI services and infrastructure

Japanese technology companies are expanding access to domestic generative AI infrastructure and models.

Sakura Internet Inc. has launched ‘Sakura AI Engine Private Edition,’ a dedicated generative AI platform for enterprises. The service allows users to host preferred AI models on the company’s GPU infrastructure and access them via API. It provides a fully managed environment covering construction, operation, monitoring, and troubleshooting. Key features include the ability to host proprietary or open-weight models in a secure, dedicated environment using domestic data centers, with a monthly fixed-rate pricing model based on GPU units. Input data is not used for AI training.

Separately, SB Intuitions Co., Ltd. has released a free beta version of the ‘Sarashina Service’ to provide access to its domestic generative AI, ‘Sarashina.’ The beta includes API functionality for six different models—covering tasks such as mini, nano, embedding, rerank, guard, and translation—as well as a web-based chat interface for dialogue and translation. The service aims to support developers and engineers in technical verification and service planning by allowing them to test how the models integrate into specific use cases.