Runway wants to turn video generation into world models and outrun Google in the AI race
Runway no longer wants to be just a service for directors and ad teams. The company is betting that the road to world models runs through video. It already works with Lionsgate and AMC, is valued at $5.3 billion, and is preparing to compete with Google not only for the video generation market but for the next platform in AI. The pitch: large language models learn from how people describe the world, while video captures motion, physics and causality directly. Runway wants to train models on that more direct data. The bet is bold but unproven. Google has video generation, its own world-model effort, and capital a startup cannot match.
AI-processed from TechCrunch; edited by Hamidun News
Runway, which the market knows for video generation tools aimed at creators and studios, is aiming at a far bigger target: building world models and competing with Google not only in video but in the next layer of AI. For the company this is not a side experiment but its new core bet.
A bet on video
While most of the industry is betting on language models, Runway's co-founders believe text alone is not enough. In their view, large language models learn from how people describe the world, not from the world itself. Video, by contrast, contains observations of motion, physics, causality and how an environment changes over time. That is why Runway treats video generation not as an end product for editing but as an intermediate step toward systems that can predict how an environment behaves.
Runway co-founder Anastasis Germanidis puts it even more bluntly: if AI is limited to human descriptions, it hits a ceiling set by knowledge already accumulated. Runway wants to train models on more direct, observable data and, in doing so, get closer to systems that do not merely draw convincing frames but actually understand how the world works. For the company, a world model is a system that can simulate an environment and forecast what happens next in a given situation.
From film to platform
Runway was founded in 2018 by three co-founders who met at NYU and originally dreamed of making films. The company's first question was simple: could AI make everyone capable of shooting a movie? After the release of its early video models, the focus shifted. The goal is now broader: to make video generation the foundation for new computing systems useful not only in film but also in robotics, science and applied simulation.
Over recent months Runway has moved this idea from slide decks into products. In December 2025 the company introduced GWM-1, its first family of world models, while continuing to develop the Gen-4.5 video model. Against that backdrop, this is not a distant research plan: the company already has a commercial business, large customers and the organizational scale to fund a riskier bet.
- valuation: $5.3 billion
- plus $40 million in annual recurring revenue added in the second quarter of 2026
- 155 employees in offices from New York to Tokyo
- partnerships with Lionsgate and AMC Networks
- a separate robotics unit launched in 2025
Runway also matters because it is not a classic Silicon Valley "lab" player. The company has no pedigree of Google alumni or Stanford graduates, and its co-founders say outright that they see being outsiders as an advantage. By their account, the early need to earn money forced the team to turn research into product faster, and the lack of Valley standardization pushed them to find their own path without ready-made recipes.
Money and competitors
Runway's main problem is not the idea but the scale of the race. Going from strong video generation to full world models takes enormous compute, a long investment horizon and access to infrastructure on the level of frontier labs. The company already works with Nvidia and CoreWeave and raised another $315 million in February 2026. Even so, that is little next to players who can afford an almost endless cycle of experiments.
The most obvious threat is Google. It already has powerful video generation and its own world-model effort, along with infrastructure, a research base and a pool of capital that a startup cannot reach. Runway also has to keep an eye on Luma, World Labs and other teams, while proving the most important thesis of all: that the path from video to a general understanding of the world is actually real. The bet is bold, but not yet proven.
"Rules are just rules somebody made up."
This remark from Runway co-founder Cristobal Valenzuela explains the company's style well. Runway does not try to play by the playbook of big AI labs and bets on speed, product discipline and an unusual starting point: creativity instead of text.
The question now is whether that will be enough when the contest moves from polished demos to a fight over compute, data and a durable advantage. They see exactly this as their main cultural advantage.
What this means
If Runway is right, video generation will stop being a niche tool for marketing and film and become a stepping stone to a more important market: systems that can model the physical world. For the industry this is a significant signal: the next big battle in AI may be fought not over the best chat answers but over machines' ability to predict and reproduce reality. That is a different market and a different logic of leadership.
- Nov 12, 2026Google opens preorders for Gemini glasses starting at $395 and blocks covert recording
- Nov 12, 2026Google brings Gemini Intelligence to Android Auto — for now, Pixel only
- Nov 12, 2026Lawsuit: Schoolgirl Spent Three Months Talking to Gemini About Suicide While Monitoring Stayed Silent for Weeks
- Nov 12, 2026Google and Adobe bring their products closer: Acrobat is coming to Gmail, Drive and Gemini
Key Questions Answered
What are the key details on A bet on video?
While most of the industry is betting on language models, Runway's co-founders believe text alone is not enough. In their view, large language models learn from how people describe the world, not from the world itself. Video, by contrast, contains observations of motion, physics, causality and how an environment changes over time. That is why Runway treats video generation not as an end product for editing but as an intermediate step toward systems that can predict how an environment behaves.
What are the key details on From film to platform?
Runway was founded in 2018 by three co-founders who met at NYU and originally dreamed of making films. The company's first question was simple: could AI make everyone capable of shooting a movie? After the release of its early video models, the focus shifted. The goal is now broader: to make video generation the foundation for new computing systems useful not only in film but also in robotics, science and applied simulation. Over recent months Runway has moved this idea from slide decks into products.
What are the key details on Money and competitors?
Runway's main problem is not the idea but the scale of the race. Going from strong video generation to full world models takes enormous compute, a long investment horizon and access to infrastructure on the level of frontier labs. The company already works with Nvidia and CoreWeave and raised another $315 million in February 2026. Even so, that is little next to players who can afford an almost endless cycle of experiments. The most obvious threat is Google.
Test Frontier LLMs & Autonomous Agents in One Workspace
Run Claude Opus 5.5, GPT-6, Gemini 3.1, and DeepSeek side-by-side in Nomad AI (nomadnet.ai) — unified model routing, deep research, and multimodal workflows under one subscription.
The AI world, distilled — once a week
Seven stories that actually mattered, hand-picked. No noise, no reposts, no press releases.
Done! Check your inbox for a confirmation.