July 22 - Reka AI, a startup spun out of former Google (GOOGL, Financial) and Meta (META, Financial) research labs, secured $110 million in Series B financing, more than tripling its valuation to over $1 billion, Bloomberg reported on Tuesday.
The Sunnyvale‑based firm, founded in 2022 by Dani Yogatama and a team of AI veterans, designs bespoke large language models for enterprise customers. Backers this round include Nvidia (NVDA, Financial) and Snowflake (SNOW, Financial), both eyeing tighter integration of Reka's LLMs into their AI ecosystems.
Prior funding in 2023 pegged Reka's worth at about $300 million, per PitchBook data. The fresh capital will fuel development of more efficient, customizable models aimed at industries from financial services to healthcare.
Snowflake's head of AI engineering, Vivek Raghunathan, said the cloud‑data platform plans to make Reka's models available through its marketplace, helping clients deploy advanced AI without managing infrastructure. “Reka's team has built capabilities rivaling major labs,” he told Bloomberg, pointing to the company's novel approach to model efficiency.
Last year, preliminary talks for Snowflake to acquire Reka fell apart as both parties decided to remain independent, according to Yogatama. He added that strategic partnerships continue, with Reka collaborating on performance tuning for Snowflake's GPU‑accelerated queries.
Nvidia did not immediately comment. Its investment signals confidence in Reka's ability to optimize inference workloads on Nvidia GPUs, positioning the startup for rapid expansion in the burgeoning enterprise AI market.