Acrab raised $350 million to build a platform for AI agents: from chips to operating systems
Singapore-based startup Acrab has raised over $350 million in total funding to build computing infrastructure for artificial intelligence agent systems. This is a significant step, given the growing demand for autonomous AI solutions capable of performing complex tasks without constant human involvement.
What will be built with these funds
The company plans to allocate the capital toward developing its own platform, research in computing systems, expanding its partner network, and entering international markets. Founded in 2024, Acrab has already established itself as a player aiming to cover the full cycle—from hardware to the software stack.
The startup is developing a computing architecture that includes:
- Proprietary AI chips
- Local execution of large language models (LLMs)
- Operating systems
- Multimodal interfaces
- AI agent management technologies
The first platform, named GΞLIX, is designed for local LLM execution in agent AI tasks. According to Acrab, the solution has already been tested in real-world conditions and is at the stage of transitioning to its first industrial deployment and mass production.
Analytical commentary
From a market perspective, raising $350 million for a startup that is only a year old is a powerful signal. Investors are clearly betting that agent AI will become the next big trend after generative models. However, the key challenge for Acrab is competition from giants like Nvidia and AMD, which already dominate the AI chip segment. Success will depend on how quickly GΞLIX can scale and offer unique advantages not available in existing solutions.