Part of the team responsible for maintaining SGLang, a popular open source tool used by companies like xAI and Cursor to accelerate the training of AI models, has transitioned to a recently launched commercial startup. The company, called RadixArk, was announced last August.
RadixArk, which was born in 2023 as SGLang in Databricks co-founder Ion Stoica’s lab at the University of California, Berkeley, was recently valued at about $400 million in a round led by Accel, according to two people familiar with the matter. TechCrunch was unable to confirm the size of the funding.
The startup previously raised angel capital from investors including Intel CEO Lip-Bu Tan, the people said.
Ying Sheng, a key contributor to SGLang and former engineer at xAI, has left Elon Musk’s AI startup to become co-founder and CEO of RadixArk, according to an announcement on LinkedIn last month. Mr. Sheng was previously a researcher at Databricks.
RadixArk’s Ying Sheng, Accel, and Lip-Bu Tan did not respond to requests for comment.
SGLang and RadixArk both focus on optimizing inference processing, allowing models to run faster and more efficiently on essentially the same hardware. Inference, along with model training, accounts for the majority of the server costs associated with AI services. As a result, using tools that optimize processes can result in huge cost savings almost immediately.
A more mature project for optimizing inference, vLLM, has also transitioned from an open source project to a startup. The newly formed company is in talks to raise more than $160 million at a valuation of about $1 billion, Forbes reported last month.
tech crunch event
san francisco
|
October 13-15, 2026
Andreessen Horowitz is leading the investment in vLLM, three people familiar with the deal told TechCrunch, but the final numbers for that investment have not yet been disclosed. Andreessen Horowitz declined to comment. In a statement to TechCrunch, vLLM co-founder Simon Moe characterized information about the round as “factually inaccurate,” but did not specify which details were incorrect.
Like SGLang, vLLM was grown in Ion Stoica’s laboratory at the University of California, Berkeley. Stoica, a professor at the University of California, Berkeley, is best known as the co-founder of Databricks and many other startups.
Brittany Walker, general partner at CRV, told TechCrunch that several large tech companies are already using vLLM to run inference workloads, and SGLang has also gained significant popularity over the past six months. Her company did not support either company.
RadixArk continues to develop SGLang as an open source AI model engine. The startup is also building Miles, a specialized framework designed for reinforcement learning. This allows companies to train AI models to become smarter over time.
Most of the company’s tools are still free, but RadixArk has started charging fees for hosting services, people familiar with the company told TechCrunch.
Startups that provide inference infrastructure to developers have seen a surge in funding in recent months, underscoring the continued importance of the inference layer to AI. Baseten recently secured $300 million at a valuation of $5 billion, the Wall Street Journal reported on Tuesday. This follows a similar move by rival Fireworks AI, which raised $250 million at a $4 billion valuation last October.
