Anastasis Germanidis
Anastasis Germanidis is a co-founder and co-CEO of Runway, an applied AI research company. He learned programming in Athens at 11, came to the United States at 18 to study neuroscience and film, and studied computer science at Wesleyan and human-computer interaction at NYU. Before co-founding Runway in 2018, he worked in backend engineering and computer-vision research. His work included Thingscoop and Uncanny Road, Soundify, Runway's Gen-2 and Gen-3 video models, research on general world models, and a 2025 diffusion vision-language model. In September 2026, he described robotics as a major application for Runway's world models and said the company was teaching them to take actions in the physical world.
- Now at
- Runway
Timeline
-
2015–2017
Thingscoop and Uncanny Road
In 2015, he builds Thingscoop, a deep-learning tool for searching and filtering videos by content. His creative projects connect software with speculative interfaces and participatory performance; in 2017, Uncanny Road uses generative adversarial networks to create a never-ending road.
-
By 2018
Wesleyan, NYU and early jobs
As an 11-year-old in Athens, he falls in love with programming; at 18, he comes to the United States to study neuroscience and film. He studies computer science at Wesleyan and completes an NYU ITP master's focused on human-computer interaction. Before Runway's 2018 founding, he works as a backend engineer at Quantcast, Chartbeat, and Zocdoc and conducts computer-vision research at IBM Research.
-
2018
Co-founding Runway
In 2018, he co-founds Runway, an applied AI research company that builds models for video and visual media.
-
2023
Soundify and Gen-2
In 2023, he co-authors Soundify, a system that matches sounds to video, synchronizes them, and adjusts spatial audio. He works on Runway's Gen-2, which generates videos from text, images, or video clips.
-
December 2023
Introducing world models
In December 2023, he starts a long-term research effort on general world models, AI systems that build internal representations of environments and simulate future events.
-
June 2024
Gen-3 Alpha
In June 2024, he introduces Gen-3 Alpha, a foundation model trained jointly on videos and images and positioned as a step toward general world models.
-
September 2025
Autoregressive-to-Diffusion model
In September 2025, he co-authors Autoregressive-to-Diffusion Vision Language Models (A2D), which adapts an existing autoregressive vision-language model for parallel diffusion decoding. The approach aims to improve the speed-quality trade-off without training a new model from scratch.
-
As of October 2026
Co-CEO of Runway
Runway names him co-CEO alongside Cristóbal Valenzuela in February 2026. In September 2026, he identifies robotics as a major application for world models and says Runway is teaching models to take actions in the world. At the Runway AI Summit in September 2026, he serves as co-CEO.
Sources and credits
Photo: We have no published photo of Anastasis Germanidis yet. Send us a free-licensed photo.
- Anastasis Germanidis | School for Poetic Computation
- Anastasis Germanidis
- Runway started by helping filmmakers — now it wants to beat Google at AI | TechCrunch
- Anastasis Germanidis | Entrepreneur | Endeavor Greece
- Soundify | Human-Computer Interaction Institute
- Gen-2: Generate novel videos with text, images or video clips
- Introducing General World Models
- Runway Gen-3 Alpha: AI Video Generation with Temporal Control
- Autoregressive-to-Diffusion Vision Language Models
- New Leaders to Support Our Next Phase of Growth
- The Kinetix Team is Joining Runway
- Runway AI Summit — San Francisco 2026