Runway Builds the First AI Router Spanning Video, Image, and Audio

Model routing has been a staple of LLM infrastructure for a while. You set a preference, the system picks the cheapest or fastest model that meets your bar, and you stop worrying about it. Runway just brought that same idea to generative media, covering video, image, and audio generation under a single router for the first time.

A catalog too large to manage by hand

The generative media model landscape has exploded. Through Runway Dev alone, developers can already access Gen-4.5, Aleph 2.0, Seedance 2, Veo 3.1, GPT Image 2, Gemini Omni Flash, ElevenLabs audio, and more. Keeping up with which model handles motion well, which one nails composition, and which one falls apart on lip sync is a full-time job most teams can't afford. Runway's Media Router encodes that expertise so developers don't have to.

Runway says this is the first model router built specifically for generative media rather than language models. The intelligence underneath it comes from Runway's in-house creative team, who have spent considerable time evaluating outputs across every media type. With LLMs, benchmarks like MMLU give you a rough signal. With video and image models, quality is far more subjective and task-dependent, which makes that human evaluation layer genuinely useful.

How the router resolves a request

Each request goes through two steps. First, the router narrows the catalog to models that are enabled in your configuration, support the capabilities your request requires, and fall under any price cap you've set. Then it selects the best remaining model according to your stated preference: cost, latency, or quality.

A router is a named, reusable configuration you create in the Developer Portal. It gets a stable config ID that you pass in your API calls. Your code references a preference, not a specific model:

import RunwayML from '@runwayml/sdk';
const client = new RunwayML({ apiKey: process.env.RUNWAYML_API_SECRET });
// Pass your router config ID instead of a model name
const task = await client.textToVideo.create({
  routerConfigId: 'your-quality-router-id',
  promptText: 'A drone shot flying over a misty mountain range at dawn',
});
// Response tells you which model ran and what it cost
console.log(task.model, task.creditsUsed);

The response includes which model ran and what it cost in credits. If no model satisfies your configuration, the request fails with a clear error. You can create as many routers as you need. A creative app might run a Draft router tuned for latency and a separate Final Export router tuned for quality, each called by its own ID.

What you can configure

The three main optimization axes are cost, latency, and quality, but the preference system goes further than that. Runway's chief product officer Anthony Maggio noted that Chinese generative media models are growing in popularity, but many businesses are uncomfortable routing production traffic through them. The router can encode that constraint directly.

That preference may become more common as the Trump administration explores restrictions on Chinese AI models. Compliance requirements are real product constraints, and the router handles them without requiring custom logic in your application.

The full set of configurable preferences:

• Optimization target — cost, latency, or quality
• Price ceiling — a maximum spend per generation so a single request never exceeds your budget
• Model allow/block lists — include or exclude specific providers or models
• Auto-upgrade — opt newly released models in automatically so routing improves without code changes

Pricing and available models

Runway Dev, launched on July 17, 2026, is a unified API platform consolidating access to AI media generation models under a single integration. It includes Runway's own models alongside third-party options. Generations cost credits at $0.01 each, billed at the standard rate of whichever model the router selects. There is no premium for using the router itself.

The cost spread the router navigates:

Runway's infrastructure bet

Runway's last dedicated frontier video model was Gen-4.5, released months ago, and its text-to-video and image-to-video models no longer lead the leaderboards. Google, ByteDance, and Alibaba have taken top spots. Rather than racing to reclaim them, Runway is positioning itself as the orchestration layer that routes across all of them.

Media Router draws directly from the routing logic Runway built for its agent product, a conversational AI creative partner launched in May that turns text prompts into fully edited multi-shot videos and marketing campaigns. Packaging that same logic for external developers extends its value without requiring Runway to win every model benchmark.

For teams building on generative media, the practical upside is straightforward: configure a router once, call it by ID, and stop updating your integration every time a model launches or pricing shifts. The router documentation is live now, and the feature is available immediately through Runway Dev.