AI-Powered
Video Editing
Training
Product
Solutions
Integrations
Free Tools
Article to Video
Turn articles into MP4
Script to Voiceover
Turn text into AI audio
Transcript to Help Article
Clean docs from transcripts
Video Trimmer
Cut and trim video clips
Subtitle Creator
Generate captions from audio
Transcript Extractor
Get text from any video
Video Watermark
Add logo or text overlay
Video Cropper
Resize and crop frames
AI Video Reframer
Auto crop with subject tracking
Video to FAQ
Turn video into Q&A
Video to Help Article
Auto-generate help docs
Video to Quiz
Generate quizzes from videos
Subtitle Translator
Translate captions instantly
Background Music
Add royalty-free tracks
Thumbnail Generator
Create click-worthy thumbs
Description Generator
SEO-optimized descriptions
AI Video Trimmer
Smart cuts, auto highlights
Video Merger
Combine multiple clips
Before & After Video
Side-by-side comparisons
Coming Next Video
PiP teaser into next clip
Video Rotator
Flip or rotate footage
Speed Changer
Speed up or slow down
Format Converter
Convert between formats
Video Compressor
Shrink file size, keep quality
Video to GIF
Turn clips into animated GIFs
Video Fade In/Out
Smooth intro and outro
Video Summary
AI-powered key takeaways
Video FAQ Generator
Extract FAQs from video
Video Report Generator
Custom report from any video
Image Annotator
Mark up screenshots
Video Annotator
Shapes, arrows & text on video
Audio Extractor
Extract audio from video
Audio Cleanup
Remove noise, level volume
Audio File Cleanup
Clean up an MP3 or WAV
Subtitle Burner
Burn captions into video
PDF to Video
Convert PDF to video
PowerPoint to Video
Convert PPT & PPTX to video
Keynote to Video
Convert Keynote to video
Presentation to Video
PDF, PPTX, or Keynote
Google Slides to Video
Turn Google Slides into video
Video Lighting
Brightness, contrast & more
Run Vidocu from your AI assistant
Connect Claude, Cursor, or any MCP client to Vidocu. Ask in plain language and your assistant turns videos into articles, translations, and courses, and records new walkthroughs with the AI Recorder.
Trusted by 20,000+ fast-moving teams
How It Works
From one URL to work done in your workspace
Connect once, approve once, then ask. Your assistant handles the tools, the order they run in, and the waiting.
- 1
Add one URL to your assistant
Add https://mcp.vidocu.ai/ as a connector in Claude, Cursor, or any client that speaks streamable HTTP. Nothing to install or host.
- 2
Approve access on Vidocu
Your client opens Vidocu's consent screen. Sign in, check what it is asking for, and approve. No API key to copy around.
- 3
Ask for what you want
Describe the outcome in plain language. The assistant picks the tools, chains them, and checks on long jobs until they finish.
- 4
Results land in your workspace
Videos, articles, translations and courses appear in Vidocu like anything you made by hand, ready to review and edit.
Connect
Hosted, remote, one line
Nothing to deploy and nothing to run locally. OAuth starts on its own the first time your assistant calls a tool.
Claude Code
claude mcp add --transport http vidocu https://mcp.vidocu.ai/Claude, Cursor, and other MCP clients
{
"mcpServers": {
"vidocu": { "url": "https://mcp.vidocu.ai/" }
}
}Client without OAuth support? Add an Authorization: Bearer header with an API key from your dashboard. Full setup, scopes, and the tool list are in the MCP documentation.
Key Capabilities
The whole platform, as tools
More than 140 tools, covering everything from a single subtitle file to a translated knowledge base and a tracked training program.
One video, every output
A single request runs analysis, subtitles, voiceover, the help article, export and translation. The assistant does not have to stitch the steps together.
Run the AI Recorder from chat
Describe a flow and the AI Recorder drives a real browser to record it. Your assistant starts the run and tells you when the video is ready.
Two-factor codes, asked in the conversation
When a recording hits a one-time code, the assistant asks you for it and passes it on. It never makes one up.
Publish, translate and ask your knowledge base
Publish a video's article to your Knowledge Center, translate it, and get answers grounded in your published articles, with citations.
Courses and training records
Build course modules from a storyboard, then ask who still has not finished a training. The planner lists people who never started, too.
Approvals and locks stay in force
Send a video for approval or lock it through a review. A locked video refuses changes from the assistant exactly as it does in the app.
Knows the price before it runs
The assistant can look up what an operation costs and your current balance, so it can tell you the price of a job before starting it.
Your permissions, nothing more
The connection acts as you. It can only do what your role in the workspace allows, and only within the scopes you approved.
Ready to hand Vidocu to your assistant?
Connect in one line
Who is it for
Every team that already works with an AI assistant
Support
Keep help articles published and translated
Turn new recordings into knowledge base articles and roll them out in every language your customers read.
Product
Record release walkthroughs without leaving chat
Ask for a walkthrough of the feature you just shipped. The AI Recorder records it and returns the video and the guide.
HR & Training
Answer training questions in one line
"Who has not finished the security module?" gets you a list, most urgent first, instead of an export to filter.
Engineering
Build Vidocu into your own agents
Use it from Claude Code or Cursor, or point your own agent at the same server. OAuth or an API key, your choice.
What Makes This Different
Skip the clicks.
Ask for the result.
Everything the Vidocu app does, your assistant can now do for you, in the order it needs, across as many videos as you ask.
| Task | Vidocu MCP | Doing it by hand |
|---|---|---|
| Who runs the steps | Your AI assistant | You do |
| Upload, article, translate, publish | One request | One screen at a time |
| Work across a whole library | Ask once | Video by video |
| "Who hasn't finished this training?" | One question | Export and filter |
| Cost of a job before you run it | The assistant tells you | Look it up yourself |
| Works from Claude, Cursor and Claude Code | Dashboard only |
Your assistant, with the whole platform in reach.
0+
Tools across the platform
Free
Works on every plan
OAuth
No API key needed
Hosted
Nothing to install
Save hours, reduce repetitive work, create more formats, and localize content for global audiences.
Frequently Asked Questions
MCP (Model Context Protocol) is an open standard that lets AI assistants use outside tools. Vidocu's MCP server gives an assistant such as Claude the Vidocu platform as a set of tools, so you can ask for an article, a translation or a recording in plain language and the assistant carries it out in your workspace.
Any MCP client that supports remote servers over streamable HTTP, including Claude, Claude Code, Cursor and Windsurf. The server is hosted at https://mcp.vidocu.ai/, so there is nothing to install or run on your machine. It is also listed on the official MCP registry as ai.vidocu/vidocu.
No. Clients that support OAuth open Vidocu's consent screen, you approve, and you are connected. For clients without OAuth support you can pass an API key from the Developers page in your dashboard instead.
Yes. The MCP server works on every plan, including Free. Operations use the same credits they would use in the app, and the assistant can check the cost before it starts a job. Features that are paid in the app stay paid through MCP: the AI Recorder needs a paid plan, and the knowledge base tools need the Knowledge Center.
Only what your account can. The connection acts as you, so your workspace role decides what it may do, and the consent screen shows which areas of Vidocu it is asking for before you approve. A locked video stays locked and an open approval round stays open.
More than 140, covering videos, projects, subtitles, voiceover, translation, articles, the AI Recorder, the Knowledge Center, courses, training, Studio, Remix, avatars, brand kits, approvals, webhooks and usage. The MCP documentation lists them.
They run in the background. The tool returns a job ID straight away and the assistant checks on it until it is done, so a long export or recording does not block the conversation.
They reach the same platform. The REST API is for code you write yourself. The MCP server is for AI assistants: each tool describes itself so the assistant knows when to use it, and you drive it by asking rather than by writing requests.
Put Vidocu inside your assistant
Create a free account, add one URL to your AI assistant, and ask for your first article.
Your knowledge is valuable. Make it usable.
Upload once. Use everywhere.
The AI Knowledge Platform. One upload becomes videos, SOPs, guides, articles, and training - in any language.
© 2026 Vidocu. All rights reserved.