Acrab has raised $350 million to build computing infrastructure for agentic AI.
Singapore-based startup Acrab has attracted over $350 million in total funding to develop computing infrastructure focused on agent-based artificial intelligence systems. These funds will be directed toward building the platform, conducting research in computing architectures, expanding its partner network, and entering international markets.
Founded in 2024, Acrab is building a full-cycle computing solutions stack: from AI chips and local deployment of large language models to operating systems, multimodal interfaces, and AI agent management technologies. The company's key product is the GΞLIX platform, designed for local execution of LLMs in the context of agent-based AI. According to Acrab, the solution has already been tested in real-world conditions and is currently in the preparation phase for its first industrial deployment and mass production.
This funding round underscores the growing market interest in specialized computing power for agent-based AI, which requires high performance and low latency. Unlike cloud-based solutions, local model deployment ensures greater data privacy and system autonomy, which is critical for corporate and industrial applications.
Expert opinion: Raising $350 million for a startup founded just a year ago is a powerful signal that the market sees potential in decentralized computing architectures for AI. However, Acrab's success will depend on its ability to scale chip production and convince major clients of the reliability of local solutions compared to cloud giants.