Most people meet AI Video Enhancer Studio through its Windows wizard: load a video, pick an enhancement preset, click Start. But the wizard is only one way to drive it. The same engine — NVIDIA RTX Video Super Resolution upscaling, RIFE frame interpolation, denoising and stabilization — also ships with a full command-line interface called aivideoenhancer, a built-in MCP server, and an installable Claude Code skill.
This post walks through all three: enhancing videos from the aivideoenhancer CLI, wiring the MCP server into Claude Desktop or Claude Code, and letting an AI agent run your enhancement pipeline from plain English. The CLI has no --json flag — for AI agents, use the MCP server instead. Full CLI and MCP reference lives in the public GitHub repository at https://github.com/contentasoftware/contentasoft-ai-video-enhancer-plugin.
The executable is called aivideoenhancer and installs together with the desktop app — same installer, same license. A default install puts it in %LOCALAPPDATA%\Programs\AIVideoEnhancerStudio\ and adds that folder to your PATH, so aivideoenhancer works from any new terminal.
Start with status to check GPU availability and license state, then analyze to inspect a file before you commit to a long enhancement run:
aivideoenhancer status
aivideoenhancer info video.mp4A single enhance command applies upscaling, denoising, and stabilization in one pass:
aivideoenhancer enhance old_tape.mp4 --output ./enhanced --upscale x4 --denoise strongPoint enhance at a folder and pick a preset — old tape restoration, anime upscaling, or low-light cleanup:
aivideoenhancer enhance ./family-videos --output ./restored --preset old_video_restorationThe engine processes files one after another on your GPU and writes the results to the output folder. Wrap it in a scheduled task and your archive converts overnight, with no per-minute billing.
MCP (Model Context Protocol) is the open standard AI assistants use to call external tools. The CLI prints JSON with --json, which suits scripts; for AI agents the MCP server is the simpler path, and you start it with one command:
aivideoenhancer serveAdd this to your claude_desktop_config.json to connect Claude Desktop:
{
"mcpServers": {
"aivideoenhancer": {
"command": "aivideoenhancer",
"args": ["serve"]
}
}
}From then on you can ask things like “enhance every MP4 in my Downloads folder with the old video restoration preset” and the agent calls the local engine directly.
If you use Claude Code, the /contenta-video-enhancer skill ships inside the AI Video Enhancer Studio plugin together with the MCP server, so one install gives the agent both. Two commands:
/plugin marketplace add contentasoftware/contentasoft-ai-video-enhancer-plugin
/plugin install contentasoft-ai-video-enhancer@contentasoft-ai-video-enhancer-pluginIn Cursor, Claude Desktop, VS Code or any other MCP client, add this server entry (needs Node.js 18 or newer):
{
"mcpServers": {
"ai-video-enhancer": { "command": "cmd", "args": ["/c", "npx", "-y", "@contentasoft/ai-video-enhancer-mcp"] }
}
}Then describe the job in your own words — “upscale this wedding folder to 4K with strong denoising” or “analyze this clip and suggest the best preset”. The skill builds the right aivideoenhancer commands and reports back.
The plugin source, the MCP server reference and the CLI documentation live in the public GitHub repository, and the launcher package is published on npm: github.com/contentasoftware/contentasoft-ai-video-enhancer-plugin · npmjs.com/package/@contentasoft/ai-video-enhancer-mcp
Most enhancement modes run on any modern GPU via DirectML or CUDA. NVIDIA RTX Video Super Resolution (VSR) — the sharpest upscaling path for web video and old tapes — requires an RTX 2000-series or newer NVIDIA card. Run aivideoenhancer status to see what your machine supports before starting a batch.
No. The CLI, the MCP server, and the Claude Code skill ship in the same installer as the desktop app and are covered by the same $129 one-time license.
Yes. The free trial covers the CLI and MCP server too. The first 5 files enhance clean; after that, output carries a small watermark until you register.
Yes. Every command takes --json and prints one JSON object with success, exitCode and outputs, and the exit codes are the same as in the other ContentaSoft command lines. AI agents can also use the MCP server, which returns typed tool results.
Download the free trial and run your first batch enhancement from the terminal in minutes, no credit card needed.
Download Free TrialWhether you prefer the wizard, a PowerShell script, or a Claude agent, AI Video Enhancer Studio runs the same local GPU pipeline. Grab the installer, try a batch from the CLI, or wire up MCP and let an agent handle the rest.