Runway launches Media Router for generative media model selection
A single API to pick the best image, video, or audio model automatically.
Runway has launched the Media Router, a new tool that automatically chooses the optimal image, video, or audio generation model for a given request based on developer-defined priorities—quality, speed, or cost. The router is available through Runway Dev, the startup's API platform released earlier this month, which provides access to both Runway's own models and an expanding roster of third-party generative media models. Chief Product Officer Anthony Maggio said the router addresses the growing fragmentation in generative media, where developers struggle to keep up with rapid model releases. The system leverages Runway’s internal creative team’s expertise in evaluating model outputs across motion, composition, and lip-syncing.
The launch positions Runway as infrastructure rather than just a model provider, with customers like Adobe, Cloudflare, ElevenLabs, Expedia, Shutterstock, and Quora already using its API. The router supports preferences such as avoiding Chinese models—a feature that may become more relevant given geopolitical tensions. It originates from routing technology Runway built for its own agent product. The move follows Runway's shift to token-based pricing and comes amid intense competition in generative video, where Runway's Gen 4.5 topped leaderboards in December. The Media Router aims to simplify developer workflows and make Runway the go-to middleware for generative media.
- Runway's Media Router auto-selects the best image, video, or audio model based on quality, speed, or cost preferences.
- Available via Runway Dev API platform, used by Adobe, Cloudflare, ElevenLabs, Expedia, Shutterstock, and Quora.
- Supports setting preferences like avoiding Chinese models; based on routing tech from Runway's agent product.
Why It Matters
Simplifies generative media model selection for developers, reducing evaluation overhead and enabling cost/quality optimization.