The AI Knowledge Platform, wired into your AI agent

Vidocu's hosted MCP server gives Claude, Cursor, Windsurf, and any MCP-aware client the full platform as tools. Turn videos, PDFs, and slides into subtitles, voiceovers, translations, and documentation from a conversation.

Read the MCP Docs Get Started Free

Connect in one line

Remote and hosted. Nothing to deploy, nothing to run locally.

claude mcp add --transport http vidocu https://mcp.vidocu.ai/

Or add it to any MCP client that speaks streamable HTTP:

{
  "mcpServers": {
    "vidocu": {
      "url": "https://mcp.vidocu.ai/"
    }
  }
}

OAuth kicks in automatically: your client opens Vidocu's consent screen and you are connected. No key management.

Works on every plan, including Free

OAuth and API keys both work on any Vidocu plan, with no credit card. An agent can connect and produce real output today, and clients without OAuth support can use an API key instead.

21 tools, the whole platform

Everything the Vidocu API can do, exposed as MCP tools. Heavy operations run async with job polling, so agents can fire work and check back.

Videos

upload_videoget_upload_urllist_videosget_video

Projects

create_projectlist_projectsget_projectupdate_projectdelete_project

Processing

process_videoanalyze_videoget_subtitlestranslate_videogenerate_voiceovergenerate_articleexport_video

Micro-tools

list_toolsget_toolexecute_tool

Jobs & usage

get_job_statusget_usage

What agents do with it

Recording to knowledge, hands-free

An agent uploads a screen recording and comes back with subtitles, a voiceover, and a step-by-step article with screenshots. One conversation, every output.

Translate a whole library

Point an agent at your tutorial library and have it translate subtitles and voiceover into 65+ languages, batch by batch, while you do something else.

Docs that keep themselves current

Wire Vidocu into an agent workflow that re-processes updated recordings and regenerates the written guide whenever your product changes.

And when there is nothing to upload yet

Vidocu's AI Recorder (beta) creates the recording itself: describe a workflow and an AI agent opens a real browser, performs the steps, and records itself doing it, returning video, step screenshots, and a written article from one run.

AI Recorder is not yet exposed through the MCP server or API. Recordings created in the app land in the same projects these tools operate on, and MCP support is on the roadmap.

Everything you need

MCP Documentation GitHub REST API

Listed on the official MCP registry as ai.vidocu/vidocu

AI Recorder

AI Subtitles

AI Voiceover

Video Translation

AI Documentation

AI Avatars

Knowledge Center

Remix

Studio

Video Editor

Zoom & Pan

Elements & Annotations

Background Music

Presentation Slides

Watermark

API

Video to Documentation

Video to SOP

Help Article Generator

AI Knowledge Base Generator

AI Video Documentation

Video to Blog Post