Runway (company)
| Company type | Private |
|---|---|
| Industry | Artificial intelligence, Technology |
| Founded | 2018 |
| Founder | Cristóbal Valenzuela, Alejandro Matamala, Anastasis Germanidis |
| Headquarters | New York City , United States |
Area served | Worldwide |
| Products | Generative AI models (Gen-1, Gen-2, Gen-3 Alpha, Gen-4), Stable Diffusion, AI video and image generation tools, iOS app, API platform |
Runway AI, Inc.—also referred to as Runway or RunwayML—is an American firm headquartered in New York City that specializes in generative artificial intelligence research and technology development. The company is dedicated to creating tools and models for producing videos, images, and various multimedia content. Its leading commercial AI models for text-to-video and video generation include Gen-1, Gen-2, Gen-3 Alpha, and Gen-4.
Runway’s AI solutions have been utilized in prominent motion pictures such as Everything Everywhere All at Once, music videos for artists including A$AP Rocky, Kanye West, Brockhampton, and The Dandy Warhols, as well as in television post-production for programs like The Late Show with Stephen Colbert and Top Gear.
History
Runway was established in 2018 by Chilean entrepreneurs Cristóbal Valenzuela and Alejandro Matamala alongside Greek co-founder Anastasis Germanidis, who met while attending New York University’s Tisch School of the Arts ITP program. The startup initially secured $2 million in funding during 2018 to build a platform aimed at deploying machine learning models within multimedia contexts.
In December 2020, the company completed a Series A funding round raising $8.5 million, followed by a $35 million Series B round in December 2021.
August 2022 saw Runway jointly launch an advanced iteration of the Latent Diffusion Model known as Stable Diffusion, developed in partnership with the CompVis Group at Ludwig Maximilian University of Munich and supported computationally by Stability AI.
On December 21, 2022, Runway raised $50 million through a Series C financing round. This was further increased by $141 million in June 2023, elevating the company’s valuation to $1.5 billion. Investors included Google, Nvidia, and Salesforce, with funds directed toward developing core multimodal AI models tailored for film and video content creation.
February 2023 marked the introduction of Gen-1 and Gen-2—the first commercially available foundational models enabling video-to-video and text-to-video generation via a web-based platform.
By June 2023, Time magazine honored Runway as one of the world’s 100 Most Influential Companies.
On April 3, 2025, Runway completed a funding round led by General Atlantic that raised $308 million, pushing the company’s valuation beyond $3 billion.
Services and technologies
Runway focuses on generative AI technologies applied to video production, media creation, and digital art. The company develops proprietary foundational models designed for use by professionals in filmmaking, post-production, advertising, editing, and visual effects. Additionally, it provides an iOS application aimed at consumer users.
Its suite of products is available through a web-based platform as well as an API delivered as a managed service.
Stable Diffusion
Stable Diffusion is an open-source deep learning text-to-image model launched in 2022. It originates from the research paper High-Resolution Image Synthesis with Latent Diffusion Models authored by Runway in collaboration with the CompVis Group at Ludwig Maximilian University of Munich. The model is primarily utilized for generating images based on textual input.
Gen-1
Released in February 2023, Gen-1 is a video-to-video generative AI framework that creates new video content by transferring the style or composition specified by an image or text prompt onto the structural elements of an existing source video. This technology was developed by Runway based on the research paper Structure and Content-Guided Video Synthesis with Diffusion Models and represents a significant advancement in generative AI-driven video production.
Gen-2
Gen-2 represents a multimodal artificial intelligence platform capable of creating original videos from inputs such as text, images, or video segments. Expanding upon the foundation established by Gen-1, it introduces video generation conditioned on textual prompts and ranks among the earliest commercially accessible text-to-video models.
Gen-3 Alpha
Gen-3 Alpha marks the inaugural release in a new generation of models developed by Runway, leveraging sophisticated infrastructure designed for extensive multimodal training. This iteration offers marked enhancements in visual fidelity, temporal coherence, and motion realism compared to Gen-2, moving closer to the concept of General World Models.
The dataset used to train Gen-3 comprised thousands of YouTube videos alongside allegedly unauthorized film content. According to a former staff member, the company undertook a coordinated initiative to catalog videos into spreadsheets before downloading them using youtube-dl via proxy servers to circumvent YouTube’s access limitations. Evaluations demonstrated the model's ability to produce videos that emulate the distinctive styles of particular YouTubers.
Gen-4
Launched in March 2025, Gen-4 stands as Runway's most sophisticated AI video generation system to date. It excels at maintaining consistent representations of characters, objects, and settings throughout multiple scenes by utilizing both reference imagery and textual instructions, thereby offering enhanced creative precision and output quality.
AI Film Festival
Runway organizes an annual AI Film Festival hosted in both Los Angeles and New York City, featuring films created entirely through AI-driven methods. This event underscores the company's pivotal contribution to the evolution of AI-powered media production.