Create AI videos (voiceover, scenes, captions, music) and post them to TikTok, Reels, Shorts and X.
One address for any MCP client. You sign in at app.makeaivideo.ai once, after adding it.
At most: moves money, runs code or commands, deletes data, or sends messages for you.
Odel charges nothing per call. The publisher may charge for its own service.
| Access to | Read | Change |
|---|---|---|
| Read | ||
| Write | ||
| Ai | Not stated: ai | |
From the sign-in at app.makeaivideo.ai, as the endpoint publishes it. You approve it there.
Odel could not list its tools without signing in.
Let Claude, ChatGPT, Cursor and any MCP client make finished AI videos and post them. The MakeAIVideo MCP server gives your AI assistant 44 tools to write a script, generate a video with AI voiceover, AI or stock scenes, captions and music, hand back the MP4, and post it (now or scheduled) to your connected TikTok, Instagram Reels, YouTube Shorts, Facebook Pages, LinkedIn, X, Threads, Pinterest, Bluesky, Telegram and Discord accounts.
Ask your assistant things like:
https://mcp.makeaivideo.aiStreamable HTTP with OAuth: add it as a custom connector and sign in, no key to copy.
https://mcp.makeaivideo.aihttps://mcp.makeaivideo.aiclaude mcp add --transport http makeaivideo https://mcp.makeaivideo.ai
~/.cursor/mcp.json) / VS Code (.vscode/mcp.json):
{ "mcpServers": { "makeaivideo": { "url": "https://mcp.makeaivideo.ai" } } }
Prefer an API key? Send Authorization: Bearer mav_... (create one in the developer settings).
For clients that only run local commands. It bridges to the hosted server, so every tool stays in sync.
{
"mcpServers": {
"makeaivideo": {
"command": "npx",
"args": ["-y", "@makeaivideo/mcp"],
"env": { "MAKEAIVIDEO_API_KEY": "mav_your_key" }
}
}
}
Works with Claude Desktop, Cursor, Windsurf, Cline, Zed, Continue and any stdio MCP client. Node 18+.
Or run it with Docker (the Dockerfile in this repo): docker build -t makeaivideo-mcp . && docker run -i --rm -e MAKEAIVIDEO_API_KEY=mav_your_key makeaivideo-mcp.
Without an API key the local server still starts and lists its tools (from the bundled tools.json snapshot), so clients can inspect it; every tool call then asks for a key. With a key, the tool list comes live from the hosted server.
| Area | Tools |
|---|---|
| Create videos | create_explainer_video, create_listicle_video, create_story_video, create_ugc_video, create_demo_video, create_article_video, create_spokesperson_video, create_video_from_script |
| Plan | estimate_video_cost, write_script, structure_text, fetch_article, generate_ideas, enhance_prompt |
| Track and deliver | get_video_status, list_my_videos, get_download_url, get_captions, create_share_link, export_video |
| Edit | update_video, generate_video, regenerate_scene, regenerate_voice, cancel_video, delete_video |
| Characters, voices, music | list_characters, create_character, generate_character_portraits, generate_voice_previews, list_voices, preview_voice, generate_music, list_music, list_templates |
| Post to social | list_social_accounts, get_connect_link, publish_video, get_publish_status, list_posts |
| Account and automation | whoami, get_credits, create_webhook, list_webhooks |
Renders take minutes: the assistant creates the video, then polls get_video_status (it respects poll_after_seconds) and returns the download link.
publish_video posts a finished video to one or more connected accounts, now or at a scheduled time, with one caption or a caption per platform. Supported: TikTok, Instagram (as a Reel), YouTube (a 9:16 video as a Short, 16:9 as a regular video), Facebook (Pages only), LinkedIn, X (videos up to 140 seconds), Threads, Pinterest, Bluesky, Telegram and Discord.
A person must connect each account once: get_connect_link returns a short-lived link that the user opens in a browser to approve access (the assistant cannot complete it). Bluesky, Telegram and Discord connect in the web app. Then list_social_accounts returns the account ids, publish_video posts, and get_publish_status returns the live post URLs.
Full reference: makeaivideo.ai/docs/mcp.
| Variable | Default | Purpose |
|---|---|---|
MAKEAIVIDEO_API_KEY | (required) | Your mav_ API key |
MAKEAIVIDEO_MCP_URL | https://mcp.makeaivideo.ai | Override the remote endpoint |
MakeAIVideo turns almost any starting point into a finished short-form video. Start from a single idea with prompt to video, or bring your own words with script to video and blog to video, which narrates a blog post you paste in. To teach a topic, AI explainer videos add voiceover, scenes, captions and music in one pass.
For a presenter on screen, AI talking avatar videos lip-sync a presenter you pick, or one made from your own reference photo, to a script. AI spokesperson videos suit corporate messaging you want to re-render when details change, and AI UGC video ads put an AI spokesperson on your ad script for Meta and TikTok. Character swap replaces the person in a video you upload while the motion and audio stay as filmed.
From a still image, image to video turns product shots, landscapes and album art into a moving clip, and animate a photo brings portraits, pets and old family photos to life.
Every video is sized for short-form platforms: use the TikTok video generator, the Instagram Reels generator or the AI YouTube Shorts generator, or run a faceless YouTube channel without filming. When a video is ready, auto-post to social media posts or schedules it to TikTok, Instagram, YouTube, Facebook Pages, LinkedIn, X, Threads, Pinterest, Bluesky, Telegram and Discord. Plans and credits are on the MakeAIVideo pricing page, and AI assistants connect through the MakeAIVideo MCP server setup guide.
From an AI assistant, the create_*_video tools cover explainer, listicle, story, UGC, demo, article and spokesperson videos plus your own scripts, and publish_video posts the result.
What is the MakeAIVideo MCP server? A Model Context Protocol server that lets AI assistants create finished short-form AI videos with MakeAIVideo.
Does it cost anything? MakeAIVideo is a paid product with a 7-day trial (card required). The MCP server has no separate charge: videos use credits from your MakeAIVideo plan, and estimate_video_cost quotes before anything is spent. See pricing.
Can it post to social media? Yes. After you connect an account once, the assistant can post a finished video to TikTok, Instagram, YouTube, Facebook Pages, LinkedIn, X, Threads, Pinterest, Bluesky, Telegram or Discord, or schedule it.
Which AI assistants work? Anything that speaks MCP: Claude (web, desktop, Code), ChatGPT, Cursor, VS Code, Windsurf, Cline, Zed and more. See AI agents.
Is there a REST API too? Yes: API reference and the TypeScript SDK.
MakeAIVideo is an AI video generator that turns a brief, a prompt or your own script into a finished, captioned short-form video: script, AI voiceover, AI-generated or stock scenes, captions and music, exported as an MP4 ready for TikTok, Instagram Reels and YouTube Shorts.
https://mcp.makeaivideo.ai| Package | What it is |
|---|---|
@makeaivideo/sdk | TypeScript / JavaScript SDK for the REST API |
@makeaivideo/mcp | MCP server for Claude, ChatGPT, Cursor and other AI assistants |
@makeaivideo/cli | Command line tool: brief in, MP4 out |
MIT © MakeAIVideo