Singapore-based neocloud Aolani has announced the launch of the Aolani Token Factory, a managed inference platform that allows organizations to deploy and scale AI models on a pay-per-token basis. This move positions Aolani as the first Singapore-founded neocloud to offer production-grade, managed inference at scale, addressing the growing demand for efficient AI infrastructure as companies across Asia invest heavily in AI-driven business outcomes.
The Token Factory eliminates the need for clients to provision or manage GPU infrastructure, instead offering a metered service where customers pre-purchase credits and pay based on token consumption. Aolani handles the entire inference stack, including GPU capacity allocation, model serving, orchestration, scheduling, and workload optimization. This approach enables companies to scale consumption without the capital-intensive burden of maintaining their own hardware.
At launch, the platform supports leading open-source models such as DeepSeek, GLM, Kimi, and Qwen, with plans to expand the catalog based on demand. Customers can also deploy custom models through OpenAI-compatible APIs. For enterprises with stringent compliance and data residency requirements, Aolani offers dedicated capacity and data isolation options.
The service is tailored for three primary use cases: AI agents for high-volume inference and workflow automation, enterprise applications like internal copilots and knowledge assistants, and coding agents for code generation and review. This versatility makes the Token Factory suitable for a wide range of AI-native companies and traditional enterprises looking to integrate AI into their operations.
Sea Xu, Applied AI Research Lead at Aolani, emphasized the platform's design for rapid model adaptation. "We built the platform to ensure our customers can get access quickly as new models emerge," Xu said. "As Southeast Asia's AI ecosystem evolves, our goal is to ensure infrastructure keeps pace."
Nicholas Chia, CEO of Aolani, highlighted the strategic significance of the launch. "Fast-moving AI natives want to build and ship products flexibly, without managing GPU fleets," Chia said. "With competitive per-token pricing and a fully managed stack, companies can go from model selection to production without the operational complexity of self-managed infrastructure."
The Aolani Token Factory is now available for interested organizations to register. For more information, visit the Token Factory page.
