Cursor gets a Runway plugin, days after Codex got the same one

Runway has released a plugin for Cursor, the AI code editor. In an announcement on X, the company said one plugin install “gives Cursor everything it needs to build Runway’s API into your app.”

The post, published on November 2, is the second of its kind in a week. The same offer reached Codex on October 28, as we reported in our earlier coverage.

One install, then Cursor writes the integration

Runway’s description is short. A developer installs the plugin, tells Cursor what the app should generate, and Cursor builds the connection to Runway’s API. The post says the user describes the generation task and “let Cursor build the integration for you.” It ends with a call to get started through a link, which was cut off in the text we could read.

The post does not say what the plugin contains, which Cursor versions or plans support it, or whether it costs anything beyond normal Runway API usage. It also gives no list of supported models. Those details are probably on Runway’s developer pages, which we could not confirm for this article. Runway’s claim that Cursor can build the integration is also the company’s own and has not been independently verified.

In practice, a developer would still need a Runway developer account and an API key, since the plugin connects Cursor to Runway’s API. Our earlier Codex article says keys are issued at dev.runwayml.com.

A pattern: The same pitch across coding agents

The Cursor announcement is nearly identical in wording to the Codex one, with only the tool’s name changed. That points to a distribution strategy rather than a one-off feature. Runway appears to be putting its API where developers already write code, so that an agent, not a person reading documentation, does the integration work.

Our earlier coverage shows this effort predates the plugins. According to that article, Runway launched a Runway Dev MCP server on September 2, 2026, at https://dev.runwayml.com/mcp. It connects a developer account to Claude, ChatGPT, Codex and Cursor. Through it, agents can compare models by price and capability, pull a model’s request schema, set up a Model Router that optimizes for cost, latency or quality, and diagnose failed generations.

The same article lists guardrails Runway cited for that service: spend management, per-model cost visibility, built-in content moderation and IP indemnification, along with 99.9% uptime and SOC 2 Type II compliance. Whether the Cursor plugin packages these capabilities, or only adds a faster install path, is not stated in the new post.

The plugins also arrive amid a run of Runway developer-facing moves. According to our earlier article, Runway MCP reached Grok Bot on October 19. Around the same time Runway promoted App Studio, which turns a plain-language description into a working tool powered by Runway’s models, and Runway Ads, according to the company’s own posts.

What changes for developers who use Cursor

For developers who already work in Cursor, the practical change is less setup. Instead of reading API documentation and writing the calls themselves, they can hand the job to the editor’s agent, with the plugin supplying Runway-specific context. This suits teams adding video or image generation to an existing product, where the work is mostly wiring and not research.

Several questions remain open. The post does not say how well the agent handles Runway’s different models, how generation costs are controlled while an agent is writing and testing code, or whether the plugin works in Cursor’s agent mode only. Developers will want to check generation spend while an agent iterates, since each test call to a video model may be billed. Pricing for the plugin itself was not disclosed.

Runway has not said whether more coding tools will follow. With Codex and Cursor covered in five days, and Claude, ChatGPT, Codex and Cursor already reachable through the MCP server, the company has the main AI coding environments in its sights.