# Runway ML MCP for AI Agents AI Agent Connect > Runway ML lets your AI client generate high-quality video clips from text or images. It connects to Runway's advanced video generation models, giving you the ability to animate still photos, create smooth transitions between frames, and manage your entire video production queue directly from your chat interface. ## Overview - **Category:** ai-frontier - **Price:** Free - **Endpoint:** https://edge.vinkius.com/vk_preview_fxPMozx8tTYwy0f2CIAyhsmnGRvf4U3doJGXFGQX/ai-agent-connect - **Tags:** video-generation, generative-ai, media-production, creative-tools, animation, model-inference ## Description You can now turn your text prompts into high-fidelity video clips without leaving your current workspace. This Connector connects your agent to the Runway ML ecosystem, making it possible to generate cinematic scenes, animate reference images, and create smooth interpolations between keyframes. Instead of jumping between different web interfaces and manual settings, you just tell your AI what you want to see. It handles the heavy lifting of communicating with the high-quality video models. You can also keep tabs on your active renders, check your organization's credit usage, and cancel jobs that aren't hitting the mark. It's a massive leap forward for anyone who needs to move from a concept to a moving visual quickly. By adding this to your Vinkius setup, you're essentially giving your AI a professional video production suite. You can prototype commercial b-roll, create stylized social content, or build out complex storyboards just by describing the motion and style you're aiming for. This means you spend less time fighting with sliders and more time refining your creative vision. Whether you need a realistic cinematic shot or a stylized animation, you can iterate fast by describing the lighting, camera movement, and subject details in plain English. It turns your chat window into a production hub where your ideas become moving assets in just a few prompts. ## Tools ### get_organization View your organization's current credit balance and details. It helps you monitor your spending across the account. ### get_task Check if a video render is finished. This provides the final output URL once the status shows as successful. ### image_text_to_video Create a video from a reference photo and a text description. It's perfect for guiding the specific motion of a still image. ### interpolate Make a smooth transition between two different images. It creates a fluid motion clip that blends the two keyframes. ### cancel_task Stop a video that is currently being rendered. Use this to save credits if you realize the prompt isn't working. ### gen3_turbo Create a quick five-second video using the Gen-3 Alpha Turbo model. It's built for speed and high-quality results. ### gen4_turbo Generate high-quality video clips using the Gen-4 Turbo model. This is the top-tier option for realistic visuals. ### image_to_video Turn a single image into a video clip. You just need to provide the image URL, the model choice, and the duration. ### list_tasks See a list of your most recent video generation jobs. This helps you keep track of everything you've submitted lately. ### text_to_video Create a video based solely on a written prompt. You can choose the model and the length of the clip. ## Prompt Examples **Prompt:** ``` Can you make a 5 second video of a cyberpunk city in the rain? ``` **Response:** ``` I've submitted that request to the Gen-3 Turbo model. Here is the status of your generation: - **Task ID**: `rw_task_5521` - **Status**: Processing - **Estimated Time**: ~1 minute I'll let you know as soon as the video is ready for you to view. ``` **Prompt:** ``` I have a photo of a mountain. Can you make it look like a time-lapse of clouds moving over it? ``` **Response:** ``` I can certainly do that. Please provide the URL for the mountain photo, and I'll use the image-to-video tool to animate the clouds for you. Once you send the link, I'll start the render. ``` **Prompt:** ``` How many credits do we have left in our organization? ``` **Response:** ``` I've checked your account details. Here is your current status: - **Organization**: Creative Studio Pro - **Remaining Credits**: 450 - **Recent Activity**: 12 tasks completed in the last 24 hours. ``` ## Capabilities ### Generate video from text prompts Create entirely new scenes using specific high-fidelity models. ### Animate still images Turn a static photo into a moving video clip with just a few words. ### Blend two images together Create smooth motion between two different keyframes using interpolation. ### Monitor production tasks See the status of your current renders and get the final output links. ### Manage organization credits View your current credit balance and usage limits to stay on budget. ### Cancel active renders Stop a generation mid-process to save on your credit usage. ## Use Cases ### Rapid social media b-roll generation A social media manager needs 10 different b-roll clips for a travel brand. They ask the agent to generate 10 unique sunset scenes using gen3_turbo and then use list_tasks to monitor the progress. ### Visualizing character movement A creative director wants to see how a character would move. They provide a character image and use image_text_to_video to describe a slow walk toward the camera to see the motion instantly. ### Seamless location transitions A filmmaker needs a transition between two locations. They use interpolate to blend a shot of a forest into a shot of a desert, creating a seamless visual bridge. ### Commercial prototyping A marketing team wants to prototype a commercial. They use gen4_turbo to create high-fidelity cinematic shots of a futuristic city based on a script, then use get_task to grab the final links for a presentation. ## Benefits - Save time on production by using gen3_turbo to generate quick five-second clips for social media. - Create complex transitions between frames using interpolate to avoid manual frame-by-frame animation. - Keep your project on budget by using cancel_task to stop renders that aren't meeting your creative vision. - Streamline your creative workflow by using image_text_to_video to turn reference photos into dynamic scenes. - Stay organized with list_tasks to track multiple video generations happening simultaneously across your account. - Monitor your spending and organization limits in real-time with get_organization to manage your production costs. ## How It Works The bottom line is you turn text or images into professional video assets without ever leaving your chat window. 1. Connect your Runway ML API keys to the Connector in your client settings. 2. Describe the video you want to create or provide a link for a reference image. 3. Get a task ID and wait for your AI to provide the final video link once it's finished. ## Frequently Asked Questions **Can the Runway ML MCP generate videos from my own photos?** Yes, you can. By providing a link to your image, your AI agent can animate it into a video clip or use it as a reference to guide specific movements. **Does this Connector support the newest Runway models?** It supports the Gen-3 Alpha and Gen-4 Turbo models, which are the current standards for high-quality video generation. **How do I manage my video production budget with this?** The Connector includes tools to check your organization's credit balance and usage. You can also cancel tasks that aren't working out to save on your costs. **Can I create smooth transitions between two different images?** Yes, the interpolation tool is specifically designed to blend two keyframe images into one smooth, fluid motion clip. **Is there a way to see all my recent video tasks at once?** You can ask your agent to list your recent tasks. It will scan your history and show you which ones succeeded and provide the links for the finished videos. **How fast is the video generation?** The Gen-3 Turbo model is optimized for speed, typically generating five-second clips very quickly for fast iteration. **How can I preview the output without fully committing my prompt?** Video generation actions using the Runway engines are directly invoked through asynchronous background tasks. Instead of an exact real-time preview, you can start small 5-second generations or iterate closely on your prompt texts until they produce the final aesthetic you want. **Does this integration process video rendering in real-time?** No. Video rendering is asynchronous. When you submit a prompt or an image, the AI will dispatch a background task to Runway's servers and receive a 'task_id'. You will need to ask the AI to verify the status of this task later, usually after a few minutes, to fetch the final video URL. **Can I use Gen-4 models for free?** No. Runway requires credits for all generations, and Gen-4 Turbo consumes significantly more credits than earlier models or standard actions. You must maintain sufficient active credits or a paid subscription on your Runway account for the AI to successfully execute the rendering jobs.