Runway MCP Brings Video Generation Directly Inside ChatGPT and Claude

Runway just showed up inside ChatGPT, and this time it is not a workaround or a third-party automation. Runway unveiled Runway MCP, allowing users to access its video creation features directly inside chat-based AI such as ChatGPT and Claude. The integration is now live in the ChatGPT app directory, meaning the tab-switching era of prompting a chatbot for ideas and then jumping to a separate video tool is officially optional.

What is MCP and why does it matter here?

MCP stands for Model Context Protocol, an open standard that lets AI assistants call out to external services mid-conversation. Think of it as a universal plugin layer for agents. MCP is a method that connects AI and external services, allowing other services' functions to be called within a chat interface. Runway is using this protocol to embed its full generation stack inside any compatible agent, not just ChatGPT.

Users can now connect Runway directly into Claude, ChatGPT, Cursor, Replit and more, to generate images and videos with models like Gen-4.5, Seedance 2.0, GPT Images 2.0, Kling 3.0, Veo 3.1, Nano Banana Pro and more. That is a remarkably broad model roster for a single integration, covering Runway's own Gen-4.5 alongside third-party models the company has aggregated under its API.

The workflow that changes

When integrated with ChatGPT, users can set video style and scene composition while talking with the AI, then proceed straight to video generation. Users can also use images entered during the chat to immediately create videos with Runway's video generation model, linking text conversation and visual input in a single workflow.

The kinds of tasks this unlocks in a single session:

• Describe a product, attach a reference image, and generate a polished marketing video without leaving the chat
• Write dialogue for characters and have Runway produce a multi-shot ad spot
• Paste a product URL and ask the agent to create a hero video for a landing page
• Generate image variants for A/B testing while iterating on copy in the same thread

Instead of moving to a separate platform to enter prompts, a flow in which users organize ideas during a chat, attach reference images, and immediately proceed to video generation could become common.

Getting connected

Setup is deliberately frictionless. Users search for Runway in the Apps menu on the ChatGPT sidebar, select connect, and log in with a Runway account. They can then call Runway functions directly within ChatGPT to carry out video generation tasks. In a supported agent, users simply add a custom connector by entering the Runway MCP endpoint and signing in with their existing Runway account. No separate API key is required.

On the billing side, MCP uses your Runway credits, just like the Runway app. The cost depends on the model, resolution, and other settings, and you can manage everything from your Runway account. All outputs are automatically saved to the user's Runway library, keeping generated assets accessible for editing or export later.

The platform play behind the announcement

This is not just a convenience feature. The change aligns with a shift in competition among generative AI services from single-model performance to workflow connectivity. Runway moved beyond a standalone video generation service and chose to bring its functions into chat interfaces where users already spend time.

The emphasis has shifted from building ever-larger standalone models to creating seamless, agent-driven workflows where models can access specialized tools and services without friction. Intelligence alone is no longer enough; the real advantage now lies in how easily those models can orchestrate tasks across creative, technical, and production pipelines. Runway is betting that being callable from anywhere is worth more than owning the interface.

The ChatGPT app directory itself is a significant distribution channel. ChatGPT now has 800 million weekly active users, surpassing even the Apple App Store's 650 million. The ChatGPT App Directory is a searchable, in-app marketplace where users can discover and connect third-party apps directly inside ChatGPT. Launched in December 2025, it replaced the earlier GPT Store model with full SDK-powered integrations that can perform tasks, access data, and trigger actions within a conversation.

Who wins, who has to adapt

Content teams and solo creators get the clearest immediate win. The ability to go from a creative brief to a finished video clip inside a single ChatGPT conversation removes a meaningful amount of friction from production workflows. Developers building in Cursor can now generate supporting visuals without breaking their coding session.

The harder question is what this means for Runway's competitors. The key to Runway MCP is that it places video generation inside other AI workflows rather than keeping it as a standalone service. Any video generation tool that stays siloed behind its own interface is now competing against one that shows up wherever the user already is. That is a distribution advantage that compounds over time.

There are real tradeoffs to flag. Runway's content moderation policies remain relatively strict, which can block or alter prompts involving sensitive, stylized, or unconventional creative directions and reduce flexibility for certain users. Because generations consume standard Runway credits regardless of the agent used, heavy or repeated use can become costly, particularly for those on lower-tier plans where model access itself is restricted. Video and complex multi-shot generations can also involve noticeable processing times, which feels more disruptive inside a conversational flow than it does in a dedicated tool.

What comes next

Runway said the tool works across multiple MCP-supported chat AIs and AI agents, including Claude and Cursor, and that accessibility will depend on MCP support and expand as more platforms adopt it. Runway has also released an open-source MCP server on GitHub, giving developers the option to explore custom implementations or extend the functionality further.

The broader trajectory is clear: creative generation tools are becoming callable infrastructure, not destinations. Runway is positioning itself as the video and image layer that any agent can invoke, similar to how Stripe became the payments layer that any app could drop in. Whether that bet pays off depends on how sticky the ChatGPT app directory becomes as a discovery and usage surface, and whether Runway's model quality continues to justify the credit costs compared to alternatives that are closing the gap fast.