Inferact, a San Francisco, CA-based developer of a high-performance AI inference engine, has raised $150 million in a seed funding round led by Andreessen Horowitz and Lightspeed Venture Partners.
The round also saw participation from Sequoia Capital, Altimeter Capital, Redpoint Ventures, and ZhenFund.
Read More:Baseten Raises $300M in Funding
The company plans to use the funds to grow its team and complete development of its vLLM technology.
Inferact sits at the intersection of AI models and hardware. While continuing to enhance the vLLM open-source ecosystem—boosting performance and hardware compatibility—the company is also building commercial products to help enterprises run AI models more efficiently at scale.
Originally developed at UC Berkeley’s Sky Computing Lab, vLLM is now managed by the PyTorch Foundation. The project has over 2,000 contributors worldwide, supports more than 500 model architectures and 200 hardware accelerators, and is used by leading tech companies, including Meta and Google.
CEO Simon Mo said inference is set to use most of the world’s new computing capacity.
Inferact’s mission is to lower inference costs and boost speed, helping drive wider AI adoption. The funding round underscores the industry’s growing focus on inference as the next frontier of AI infrastructure.
The team includes Mo, Woosuk Kwon, Kaichao You, Roger Wang, Joseph Gonzalez, Ion Stoica and others.
About Inferact
Inferact, founded in November 2025, is led by CEO Simon Mo, a UC Berkeley PhD candidate and one of vLLM’s founding maintainers. Co-founder Kaichao You, a PhD from Tsinghua University, is also a core vLLM contributor, and the founding team includes other key vLLM contributors, including Woosuk Kwon and Joseph Gonzalez. Inferact develops and supports an open-source inference engine that helps companies run large language models (LLMs) while reducing operational costs and improving performance through advanced memory management techniques, such as PagedAttention.
Read More:Cork Raises $5.5M in Seed Funding
