Singapore-based startup Acrab has raised $350 million for agentic AI infrastructure.
Singapore-based company Acrab has closed a funding round worth over $350 million. The funds will be directed toward creating specialized computing infrastructure for agentic artificial intelligence systems — one of the most promising and capital-intensive areas in the modern AI industry.
The raised capital will allow Acrab to accelerate the development of its own platform, intensify research in high-performance computing, expand its network of strategic partnerships, and enter international markets. This is one of the largest rounds for infrastructure AI startups from Southeast Asia.
Full-cycle architecture
Founded in 2024, Acrab positions itself as a developer of end-to-end computing architecture for agentic AI. The technology portfolio includes proprietary AI chips, solutions for local deployment of large language models (LLMs), operating systems, multimodal interfaces, and AI agent management systems.
The company's first product is the GΞLIX platform, designed to handle LLM workloads in agentic AI tasks directly on the device — without mandatory reliance on cloud servers. According to Acrab, GΞLIX has already been validated in real-world scenarios and is currently at the stage of transitioning to industrial deployment and mass production.
My expert opinion: Raising such significant capital at an early stage of a startup's existence is a strong signal to the market. Agentic AI requires a fundamentally different computing paradigm compared to traditional cloud models, and Acrab is betting on infrastructure that could become the standard for the next generation of decentralized and local AI systems. The success of GΞLIX on an industrial scale could seriously shift the balance of power in this niche.