Inferact Raises $150M to Scale vLLM AI Inference Technology | Ukraine news

The developers of the open-source vLLM project announced the transition of the popular tool into the startup Inferact, which raised $150 million in a seed round at a valuation of around $800 million.

The round was led by Andreessen Horowitz and Lightspeed Venture Partners, underscoring strong investor interest in commercializing inference technologies.

Inferact’s debut mirrors the path of SGLang, known as RadixArk: according to sources, the company secured funding at a $400 million valuation with participation from Accel.

The focus of artificial intelligence on deploying models in applications, a process called inference, makes technologies like vLLM and SGLang faster and cheaper, attracting investor attention.

As in 2023, vLLM and SGLang were developed in UC Berkeley’s laboratory during the work of Ion Stoica, co-founder of Databricks.

The CEO of Inferact, Simon Mo, one of the project’s early founders, said that among the existing vLLM users are Amazon’s cloud service and a mobile shopping app.

In the future, inference is expected to remain at the core of fast and affordable deployment of large language models in real-world applications, and Inferact plans to expand its customer base and scale AI deployment solutions across environments – from cloud services to mobile devices.