RE : AI is the sovereign cloud service under Singtel of Singapore Telecommunications (Digital InfraCo). Today, they have launched AI Token - as-a-Service ( TaaS), claiming that this is the first such service in Singapore. Its aim is to help enterprises accelerate their adoption of AI while addressing key challenges such as costs, model management, and data sovereignty.

TaaS is a fully managed service that provides enterprises with a AI model, scalable infrastructure, orchestration, and governance capabilities through a flexible token-based consumption model.

As enterprises expand the application of AI across various businesses to enhance productivity and efficiency, concerns regarding rising token costs and sending sensitive business data to overseas-hosted AI models may hinder broader adoption. These challenges become even more significant as companies introduce agentic and AI. Moreover, agentic and AI may increase token consumption and raise the complexity of managing multiple models and AI capabilities.

RE : AI indicates that TaaS addresses these issues by hosting the AI model on sovereign infrastructure in Singapore, which is supported by the award-winning RE : AI sovereign AI cloud and integrated technology stack. This infrastructure encompasses AI data centers, robust land and undersea connections, as well as a Paragon orchestration platform. This enables enterprises to strike a balance between data sovereignty and cost, performance, and scalability.

TaaS enables enterprises to better control AI usage and costs through flexible subscription options. Enterprises can choose the desired amount of tokens. These tokens can be used for a variety of AI applications, ranging from internal queries and software development to agentic AI; the latter allows AI to automate work processes, conduct market research, and review documents.

Through its models-as-a-service product, TaaS enables enterprises to utilize a variety of AI models without the need to purchase and manage each one individually. Enterprises can also take advantage of TaaS's intelligent routing feature, which automatically matches workloads with the most suitable AI models to optimize costs and performance.

Applications such as software development, content creation, and voice and video services can leverage leading open-source weighting models like Qwen, Mistral, GLM, Kimi, and MiniMax, as well as some cutting-edge models when necessary, to process tasks at a lower cost. Enterprises can also integrate their own models to meet specific workloads and business requirements.

Click here to read the full press release.