AI video generation startup Runway breaks the mold of a conventional Silicon Valley venture. Founded by three creative minds — two from Chile and one from Greece — who met at NYU’s Tisch School of the Arts, Runway has built its operations in New York City without the usual backing of high-profile connections or massive seed funding.
While some might view Runway as another player in the AI landscape, its ambitions suggest it could redefine the industry. The startup is challenging the assumption that intelligence is rooted in language alone, a notion underscored by the emergence of large language models like ChatGPT from OpenAI and Claude by Anthropic. Instead, Runway’s founders are betting on a future where AI evolves beyond text, harnessing video and world models that understand real-world dynamics rather than just human language.
Co-founder and co-CEO Anastasis Germanidis emphasizes the need to train models on observational data directly from the environment. He asserts that success in AI will not come from refining language alone but from utilizing less biased, more diverse data to grasp the complexities of reality. “We’re essentially confined by our current perceptions,” Germanidis explained during an interview at Runway’s bright, welcoming headquarters near Union Square.
Runway, which debuted in 2018, has established a reputation for innovative video-generation technology, most recently with its Gen-4.5 model. Their AI tools allow users to convert text prompts into cinematic content, enhancing workflows for filmmakers and advertising agencies. Partnerships with major names like Lionsgate and AMC Networks showcase the impact of their technology, which has even made its way into award-winning films such as “Everything Everywhere All At Once.”
With a valuation of $5.3 billion, Runway reported a notable increase of $40 million in annual recurring revenue in the second quarter of 2026. If their shift towards video-based world models succeeds, it could have far-reaching implications across various fields, from entertainment to pharmacology. However, if they falter, competitive threats loom, particularly from tech giants with extensive resources like Google.
In the past six months, Runway has set its plans in motion, venturing beyond video generation. They launched their first world model in December and are poised to roll out another this year. These world models are designed to mimic environments accurately enough to predict behavior, with potential applications in interactive entertainment, robotics, and gaming, competing with similar initiatives from startups like Luma and World Labs, as well as Google's Genie model.
All players in this space are striving for a common goal: creating AI solutions for some of humanity’s most daunting challenges. This expansion diverges from Runway's original focus, but it reflects both the technology's evolution and the founders' willingness to adapt.
Germanidis envisions world models as a foundational scientific tool. By feeding a model diverse sensory data, it could closely replicate a digital version of the universe, facilitating experiments far more swiftly than conventional laboratories. He argues that condensing experimental wait times could significantly accelerate scientific progress.
Having developed its products with the idea of democratizing filmmaking at its core, Runway’s mission has matured to empowering people to become exceptional filmmakers. Following the launch of their initial video model in February 2023, the team has grown to about 155 employees across offices in New York, London, San Francisco, Seattle, Tel Aviv, and recently, Tokyo.
The founders have identified that as these models evolve, their understanding of how the world operates broadens. This, in turn, opens the door to tackle significant challenges in various sectors, including robotics and drug discovery. Runway also initiated a robotics division last year, already promising real-world implementation and testing.
Germanidis believes that the future lies in training a single model on various mediums — text, video, audio, and other sensory data — resulting in enhanced capabilities. He sees the potential for Runway’s technology to impact fields like medical research, particularly in anti-aging studies.
Runway faces a significant challenge: transitioning its prowess in video generation to the rapidly evolving arena of world models. The competition includes established entities like OpenAI and Google, which command vast financial resources and established reputations.
Kian Katanforoosh, CEO of the AI benchmarking firm Workera, noted that no one has yet demonstrated a proven connection between video intelligence and broader reasoning skills, but the possibility remains. He emphasizes that for Runway to realize its ambitions, it must secure adequate resources for large-scale computation.
Despite raising $860 million to date, including a $315 million infusion in February from partners such as AMD Ventures and Nvidia, Runway finds itself in a competitive landscape where major players like OpenAI and Google are well-resourced. Katanforoosh highlights that while resources are essential, they do not guarantee success, as seen with OpenAI's recent challenges in maintaining its video platform.
Nonetheless, he doesn't dismiss Runway's potential, pointing out successes like AI audio startup ElevenLabs, which outperforms larger incumbents without their funding. Runway’s founders share a conviction that their unconventional journey provides a strategic advantage, attributing their success to a diverse range of perspectives and a lack of dependence on standard Silicon Valley practices.
According to Michelle Kwon, Runway’s COO, the startup is not rushing to seek additional funding, even with their growing compute needs. Early investor Michael Dempsey praised Runway for its agile culture and strategic foresight fostered by its founders’ unique backgrounds.
For Cristóbal Valenzuela, co-CEO, this innovative mindset is essential. He dedicates his limited free time to exploring literature, drawing inspiration from figures like Chilean poet Nicanor Parra, whose unconventional approach resonates with Runway's ethos. Valenzuela underscores that the so-called “rules” of Silicon Valley can be disregarded, advocating for a fresh perspective to ignite innovation.
