Developers can now use Runway Media Router to automatically select the best generative model for images, video, or audio. The tool aims to simplify integration and optimize for quality, speed, or cost.
Runway has introduced a new tool designed to help developers and media companies navigate the increasingly crowded field of generative AI models. The Runway Media Router, launched through the Runway Dev platform, enables users to automatically select the most suitable image, video, or audio generation model based on specific priorities such as quality, speed, or cost.
According to Runway, this router is the first of its kind built specifically for generative media, rather than large language models. The company aims to position itself as the infrastructure layer for generative media, moving beyond its origins as an AI video startup. Through Runway Dev, customers-including Adobe, Cloudflare, ElevenLabs, Expedia, Shutterstock, and Quora-can integrate media generation directly into their products via API, eliminating the need to redirect users to Runway’s own platform.
The launch comes at a time when the number of generative media models has surged, making it challenging for developers to evaluate and choose among new releases. Runway’s chief product officer, Anthony Maggio, said the Media Router leverages an intelligence layer developed by the company’s creative team, which assesses model performance across various media types. This allows developers to set preferences not only for quality and speed but also for factors like token pricing and model origin, such as favoring American providers over Chinese ones if desired.
Runway recently shifted from unlimited subscription plans to token-based pricing, a move that drew criticism from some users. The company says the Media Router’s intelligence layer is informed by its experience evaluating outputs for its own agent product, which helps turn text prompts into edited videos and campaigns. Now, this routing technology is available for external developers to use in their own workflows.
As competition intensifies, Runway’s strategy reflects the fragmented nature of the generative media landscape. While its last major video model, Gen 4.5, launched in December and briefly led industry rankings, newer models from companies like Google, ByteDance, and Alibaba have since taken top spots. Rather than betting on a single model’s continued dominance, Runway’s Media Router is designed to adapt as the best-performing models change over time, keeping the company relevant as an orchestration layer for generative media.
Runway’s co-founder and co-CEO, Anastasis Germanidis, noted that the company has built a full stack-including a developer platform, creative tools, and an inference layer-to meet growing demand from businesses seeking to integrate generative media at every level. He emphasized that orchestration and intelligence layers are becoming increasingly important as companies build complex campaigns and multi-scene generations using these models.
Runway was founded in 2018 and has since evolved from a niche AI video startup to a recognized player in the generative media infrastructure space. The company’s client roster includes major technology and media brands, and its platform supports a growing ecosystem of third-party and proprietary models for image, video, and audio generation.
