VidXP MCP server
Commit Nov 14, 2026Lets AI clients import, index, and search videos by speech, sound, scene, action, and recurring faces with inspectable evidence.
Also known as grayhatdevelopers/vidxp
- Stars
- 59
- Tools
- —
- Downloads, 30 days
- —
- Latest version
- —
- Last commit
- Nov 14, 2026
- License
- MIT
What is VidXP MCP?
VidXP lets AI clients import, index, and search videos by speech, sound, scene, action, and recurring faces with inspectable evidence.
Summary generated from the project's README
About VidXP MCP server
Searching through video libraries by describing specific moments is what VidXP enables for developers and AI agents. This local-first engine allows users to index video collections and retrieve content using natural language queries based on spoken words, ambient sounds, visual scenes, or complex actions. It is designed to work on Windows, Linux, and Apple Silicon macOS, providing a way to keep personal or project files on your own machine.
Developers can integrate these capabilities into applications via a Python package, an HTTP API, or through the Model Context Protocol. The server supports various deployment methods including a desktop application, a command-line interface using uv, or a Docker container for home servers. Once indexed, the system provides inspectable evidence such as frames, clips, or boards to verify search results. Specific tools like vidxp search speech, vidxp search sound, and vidxp search action allow for precise retrieval of media content based on semantic meaning rather than just timestamps.
What you can do with it
- 1Find specific spoken dialogueLocate a moment in a video by typing the exact words someone said.
- 2Search for visual actionsIdentify a sequence of events by describing the movement occurring on screen.
- 3Identify recurring peopleGroup and find specific faces that appear multiple times across your video files.
- 4Retrieve audio eventsFind specific sounds like barking or traffic within your media library.
Written from the project's README. Every use case is checked against a quote from it.
Automated checks
Ran Nov 14, 2026Full report →Not verified yet. A server needs a score of 80, no serious finding, and enough data for us to check. How checks work · What to fix for more points
- Source repository linked and reachablegrayhatdevelopers/vidxp is public and reachablegithub.com/grayhatdevelopers/vidxp ↗Pass
- Package names this repository as its sourceNot checked
- Build provenance (Sigstore attestation)Not checked
- Open source license (OSI)MITspdx.org/licenses/MIT.html ↗Pass
- Commit or release in the last yearLast commit or release 2026-11-12github.com/grayhatdevelopers/vidxp ↗Pass
- OpenSSF ScorecardNot checked
- Package name not close to a popular packageNot checked
- No critical advisory in the package or its dependenciesNot checked
- No hidden instructions to the model0 tool descriptions and the README read; no override phrases, credential paths or hidden charactersPass
- Destructive tools are labelledNot checked
- Remote endpoint answers the MCP handshakeNot checked
README
View on GitHubFind the moment, not the timestamp
VidXP makes one video—or an entire collection—searchable by meaning:
- Speech search: type what you remember someone saying and jump to the matching moments.
- Scene search: describe what appeared on screen and find the closest visual matches.
- Action search: describe something that happens over several seconds.
- Actor matching: find recurring faces within a video and export a highlighted video for a selected group.
Use it to search years of family videos, add video search to an editing workflow, or let an AI agent answer questions using evidence from your own video library. Your videos can stay on your machine.
Start here
Choose the setup that fits how you want to use VidXP.
1. Desktop app
Download the installer for Windows, Apple Silicon macOS, or Linux from GitHub Releases.
Connect an existing VidXP installation or let the desktop app manage an isolated runtime for you. See the Desktop installation instructions for supported setup options.
2. CLI and local AI assistants
For commands, scripts, and local AI assistants, install uv, then run:
# Install the CPU edition
uv tool install --python 3.14 --torch-backend cpu "vidxp[local-worker,mcp]"
# Check FFmpeg, download models, and verify the installation
vidxp init
vidxp prepare
vidxp doctor
# Print the settings for a local MCP client
vidxp mcp-config
Add the browser interface with:
uv tool install --python 3.14 --torch-backend cpu \
"vidxp[local-worker,mcp,frontend]"
vidxp ui
See the installation guide for client-specific MCP configuration, the HTTP API, and remote server setup.
3. Docker for a server
Run the published all-in-one image on a home server or another single machine:
# Download the search models into the persistent volume
docker run --rm -it \
-v vidxp-data:/var/lib/vidxp \
ghcr.io/grayhatdevelopers/vidxp:latest \
vidxp prepare
# Start the browser interface
docker run --rm --init \
-p 8501:8501 \
-v vidxp-data:/var/lib/vidxp \
ghcr.io/grayhatdevelopers/vidxp:latest
For a long-lived server, pin a published version instead of latest. For a
Coolify deployment, use the published -control and -worker images with
compose.coolify.yaml—no repository build is required.
See the Coolify guide for the complete setup.
What you can do today
- Build searchable libraries from individual videos or whole collections.
- Find speech, sound events, visual scenes, and multi-frame actions by description.
- Ask grounded questions and inspect the supporting boards, frames, or clips.
- Group recurring faces and render highlighted actor overlays.
- Keep personal, client, or project libraries separate.
- Use VidXP through the desktop app, browser, CLI, MCP, or HTTP API.
A first search
The browser app guides you through importing and indexing. The same flow from the command line is:
# Add a video
vidxp media import samplevideo.mp4 --json
# Index the returned media ID
vidxp index create <media-id>
# Find a visual moment
vidxp search scene "a yellow taxi on a city street"
# Find an action or event
vidxp search action "a person opens a door and walks outside"
# Find a sound event
vidxp search sound "a dog barking over traffic noise"
# Find something that was said
vidxp search speech "the bread just came out of the oven"
Results include the source video, timestamps, match score, and the evidence
used to find the moment. Add --media-id <media-id> to search only one video.
Run vidxp --help or vidxp <command> --help for the full command reference.
For applications and AI agents
Use the Python package to add selected VidXP capabilities directly to an application, or use the HTTP API when VidXP runs as a service.
MCP lets AI clients add and index videos, search a library, ask grounded questions, and return inspectable evidence such as boards, frames, and clips. A local client can start VidXP as a program on the same computer. A hosted client connects to a deployed VidXP server.
Codex plugin and skills
VidXP is distributed as a Codex plugin through a Git marketplace hosted in this GitHub repository. It includes three reusable workflows:
- install Desktop or the CLI and connect Codex;
- ingest and index videos; and
- find moments and return inspectable evidence.
Paste this into Codex:
Add https://github.com/grayhatdevelopers/vidxp as a Git plugin marketplace, install the VidXP plugin, then use its $vidxp-install skill to set up VidXP on this computer.
VidXP Desktop can perform the same setup from its Set up in Codex button.
Compatible AI clients can show an interactive upload and evidence-review view. Clients without that interface still receive the same workflow results through ordinary MCP tools.
- Python, HTTP, and MCP installation
- Local HTTP API and MCP server
- Premiere Pro extension
- ChatGPT and Codex plugin integration
- Optional capability packages
- Coolify server setup
Downloads and storage
First setup downloads only the models needed for the capabilities you select. VidXP shows the download size and destination before it starts.
The Desktop-managed Python runtime and its selected dependencies can use approximately 3 GiB.
| Capability | Approximate model download |
|---|---|
| Speech search | 2.64 GiB |
| Sound event search | 0.94 GiB |
| Scene search | 1.43 GiB |
| Action search | 0.93 GiB |
| Actor matching | 37 MiB |
A full local Desktop setup with every search capability uses approximately 9.0 GiB. Leave additional temporary space during installation and for indexes, source videos, and exported results.
By default, the CLI and desktop app share the same VidXP data directory:
| Platform | Default location |
|---|---|
| Windows | %LOCALAPPDATA%\VidXP |
| macOS | ~/Library/Application Support/VidXP |
| Linux | ${XDG_DATA_HOME:-~/.local/share}/VidXP |
Docker keeps the same data in the vidxp-data volume shown above.
Product roadmap
The next product improvements are focused on:
- labeling actor groups and matching the same person across different videos;
- more reliable face tracking across angle, lighting, motion, and occlusion;
- connecting visible people with the dialogue they are speaking;
- better search ranking, time ranges, and natural-language questions across a whole library;
- richer previews, timelines, filters, saved searches, and result playback;
- easier organization for large personal and project video collections;
- faster indexing and supported GPU acceleration; and
- smoother desktop updates, repair, and model management.
VidXP is in beta. Feedback about search quality, actor workflows, and real video-library use cases is especially useful.
Help and project links
- Installation and troubleshooting
- Desktop development
- Coolify deployment
- Changelog
- Issue tracker
- MIT license
Contributing
Contributions are welcome. Read the contribution guide before opening a pull request.
Credits
VidXP began as a student research project by:
The research was conducted with Dr Shahab Tahzeeb at NED University of Engineering and Technology and Saad Bazaz at Grayhat.
VidXP is now built by Grayhat and maintained by community contributors.
Email: info@grayhat.studio
VidXP MCP questions
What can I search for in videos with VidXP?
You can search by spoken words, sound events, visual scenes, multi-frame actions, and recurring faces. Each result includes the source video, timestamps, match score, and supporting evidence.
How do I use VidXP with an AI client over MCP?
Install the mcp extra, then run vidxp mcp-config to print the settings for a local MCP client. MCP clients can add and index videos, search a library, and ask grounded questions.
Do my videos leave my machine?
No, VidXP is local-first and your videos can stay on your machine. A hosted client instead connects to a deployed VidXP server.
How much disk space do the search models need?
Model downloads range from about 37 MiB for actor matching to 2.64 GiB for speech search. A full local setup with every capability uses roughly 9.0 GiB.
Do I need an API key or login to use VidXP?
The README does not mention any API key, token, or OAuth login for using VidXP locally through the desktop app, CLI, MCP, or HTTP API.

