# HeyGen MCP MCP > HeyGen connects your AI client directly to a powerful video studio backend. Generate high-quality, realistic talking-head videos using digital avatars that speak in any language. Use pre-set templates for brand consistency or feed simple text prompts for quick, one-shot content creation for marketing and training. ## Overview - **Category:** ai-frontier - **Price:** Free - **Tags:** ai-avatars, video-generation, text-to-video, synthetic-media, content-creation, digital-human ## Description Need professional video content but hate spending hours recording and editing? This MCP lets you skip the studio entirely. You run full AI production workflows right through your agent interface. Just tell your client what you need—a product announcement or a compliance tutorial—and it manages everything from selecting avatars to rendering the final file. It’s about scale. Instead of manually adjusting timelines, you define parameters and let the system handle the video generation. You can create videos tailored to specific brand templates for massive campaigns or use simple text inputs to generate completely new assets on demand. If your agent client is compatible with Vinkius, this MCP gives it instant access to advanced media tools that usually require complex API calls. ## Tools ### remove_video_record Deletes a specific video record from your managed library. ### create_from_template Generates a video by automatically filling in variables within a predefined design template. ### generate_ai_video Creates a full-featured talking head video using an AI avatar and provided text. ### get_api_status Checks the overall operational status of the connected account and API key. ### get_video_progress Monitors a video generation job using its task ID, returning progress percentage and eventual download links. ### list_available_avatars Provides a list of all digital avatars available for use in videos. ### list_video_templates Retrieves available video templates, allowing you to see pre-designed structures for different content types. ### list_generated_videos Shows a list of all videos that have been previously created through the system. ### list_available_voices Retrieves a catalogue of professional voices, including language and style details. ### test_heygen_auth Verifies that the API credentials provided are correct and active. ### upload_media_asset Adds external media, like an image or audio file, to your content library for use in videos. ### generate_from_prompt Produces a complete, single-shot video based only on a simple text description. ## Prompt Examples **Prompt:** ``` Generate a video from prompt: 'A professional man introducing our new AI cloud services'. ``` **Response:** ``` Triggering Video Agent... I've started a new generation job with your prompt. The task ID is 'v123456789'. I'll monitor it for you and provide the download link as soon as it's ready. ``` **Prompt:** ``` Show me a list of all available English voices in HeyGen. ``` **Response:** ``` Fetching English voices... I found 15 high-quality options. Popular choices include 'Aria' (Female, Professional) and 'Davis' (Male, Narrative). Which one would you like to use for your next video? ``` **Prompt:** ``` Check the status of video job 'v123456789'. ``` **Response:** ``` Checking status... Video 'v123456789' is now 100% complete! You can download it here: [download_url]. The video features a professional avatar with high-fidelity synthesis. ``` ## Capabilities ### Generate Video from Text Creates a complete video using an AI avatar and text input. ### Build Videos from Templates Generates personalized videos by automatically replacing variables in established design templates. ### Manage Media Assets Uploads images, audio files, and other media to use as inputs for video creation. ### Discover Available Avatars and Voices Retrieves lists of available digital avatars and professional voices for selection. ### Track Video Jobs Checks the current status and retrieves final download links for long-running video generation tasks. ## Use Cases ### Updating Product Documentation A technical writer needs to update a compliance guide across 50 different regional manuals. Instead of recording 50 videos, they connect the MCP, use `list_video_templates` to find the right structure, and then run `create_from_template`, updating only the local regulation text for each region automatically. ### Running a Sales Campaign A sales team member has 100 prospects. They want to send a personalized video message mentioning the prospect's company name and product of choice. The agent uses `upload_media_asset` for the logo, then calls `generate_ai_video`, passing in unique variables for each client. ### Rapid Ad Testing A marketer needs 20 versions of an ad to test different headlines. They use `list_available_avatars` to pick a consistent face, and then cycle through various short prompts using `generate_from_prompt`, quickly feeding the resulting assets into their platform. ### Monitoring Large Batches A training department sends out 50 videos for review. They use `list_generated_videos` to see all job IDs, and then periodically call `get_video_progress` until every single asset is marked as complete. ## Benefits - Scale production without hiring more editors. Use `create_from_template` to generate hundreds of personalized ads by simply updating variable fields like product names or client logos. - Skip the setup time. If you just need a quick demo video, use `generate_from_prompt`. It lets your agent turn any text idea into a visible asset immediately. - Manage all inputs in one place. Before generating anything, use `upload_media_asset` to add custom images or audio clips so the avatar can interact with your brand's specific visual identity. - Never wonder if the job finished. Use `get_video_progress` whenever you need a status check on a long-running render, getting instant download links when it’s done. - Ensure consistency across campaigns. By listing assets using `list_available_avatars` and checking voices with `list_available_voices`, you keep your brand's digital presence uniform. ## How It Works The bottom line is, you tell your agent client *what* video to make, and it manages all the steps needed to generate and deliver the finished file. 1. Subscribe to this MCP and provide your API key. 2. Instruct your agent client to list available avatars or voices first. This confirms the connection is working and shows you what assets are ready. 3. Use a specific tool, like `generate_from_prompt`, giving it the text and avatar details. The system then kicks off the job and provides a task ID for later tracking. ## Frequently Asked Questions **How do I check if this MCP is connected correctly using get_api_status?** Run `get_api_status`. This tool immediately verifies your credentials and tells you if the account connection is active, saving you from wasting time on failed generation attempts. **What's the difference between generate_ai_video and generate_from_prompt?** Use `generate_from_prompt` for quick, ad-hoc videos based only on text. Use `generate_ai_video` when you need more control over variables or are using a structured template. **Can I add my own images to the video with upload_media_asset?** Yes, `upload_media_asset` lets you add any image or audio file. You then reference this asset ID when calling your main generation tools. **How do I find out what kind of videos I can make? (list_video_templates)** Calling `list_video_templates` shows you all the pre-built structures available. This helps you pick a proven format that already works for your industry. **How do I use `get_video_progress` to track a long video generation job?** You pass the unique task ID into `get_video_progress`. This tells you exactly where the video is in the queue, whether it’s 25% done or still processing. It's essential for monitoring large batches so you know when to expect the final download link. **What does `list_available_voices` show me when I need a specific language or accent?** This tool fetches a comprehensive list of every voice available in HeyGen. You can filter this data by gender, language, and tone (e.g., professional vs. casual). This lets you pick the perfect voice for your content without guesswork. **If I make a mistake or need to clean up old files, how do I use `remove_video_record`?** `remove_video_record` deletes specific video creations from your library. You just need the record's ID to run it. This keeps your asset management clean and ensures you only work with current content. **Should I always run `test_heygen_auth` before starting any video generation?** Yes, running `test_heygen_auth` verifies that your API key is active and correctly configured. It quickly confirms the connection between this MCP and HeyGen, preventing frustrating errors later. **How do I choose the right avatar for my video?** Use the `list_available_avatars` tool to browse all options. Your agent will provide the names and unique `avatar_id` for each character, which you can then use in your generation requests. **Can I generate a video without specifying an avatar or voice?** Yes! Use the `generate_from_prompt` tool (Video Agent). Simply describe what you want, and HeyGen will automatically select the best-matching avatar and voice for your content. **Is it possible to automate personalized videos for multiple customers?** Absolutely. Use the `create_from_template` tool by providing a `templateId` and a set of variables. Your AI agent can iterate through your customer list and trigger unique videos for each person.