NvidiaVideo-creating startup Luma AI is joining a growing wave of US tech companies to launch US tech operations in the UK, with major plans for a London expansion on Tuesday.
The Palo Alto-headquartered company plans to hire about 200 employees, about 40% of its workforce, at its new London base in early 2027, working across research, engineering, partnerships and strategic development.
The extension is two weeks, reports Luma $900 million funding round The state investment fund financed by Saudi Arabia is Kazakhstan’s Ilgal, which estimates it to be worth up to 4 billion dollars. Launch previously from Nvidia.
Luma “World Models” are being built, consisting of video, audio and images, text, with text and which large language models (LLMS) (LLMS) and opai’s chat and Google Using Gemini.
The startup currently markets to the marketing, advertising, media and entertainment sectors, with video models that it sells as part of an application programming interface (API) and content creation suite.
“With Series C, the growth of global computing infrastructure and going forward, we have the capital and opportunity to build world-class AI everywhere,” said Amit Jain, CEO and founder of Luma Ai, everywhere. “Launching in Europe and the Middle East is the logical next step in putting the power directly in the hands of global events, agencies and brands.
The U.K. is the U.K. because it’s accessible to Darren, Jane told CNBC.
“London has some people involved in research who have given to universities here and institutions like Deepning. “We also see London as an entry point to the European market.”
The image generated by the AI is generated by Luma’s Ray3 model (Luma AI).
Forward AI
Luma is the latest in the North American wave as they double down in the UK and Europe as they capitalize on talent pools and revenue opportunities.
In November, the San Francisco-based anthropic will open offices in Paris and Munich, and launch a rental spree in London and Dublin. The Canadian AI startup Mama said it would open a Paris office to serve as its EMEA headquarters in September, while Assoida announced a new office in Munich in February.
Between those two World models Although not designed as LLMS, some researchers like them, if not more. Trying to achieve Artificial General Intelligence (AGI).
“These types of visual models are about a year right now and a year behind language models,” Jain said.
But the world models will be a “Natural Interface” for AI, that is for everyday use, it will use time ever, that will watch a video every day.
Tech giants incl Goont, Meta and Nvidia develops global models in a number of cases.
Luma, in September, released its latest model, which Jane told CNBC benchmarks is superior to the oplays sora and the Google Veo 3 at similar levels.