Which integrations are available for this server?

Provides subject segmentation using MediaPipe.

1. Click on "Deploy Server". 2. Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state. 3. In the chat, type @ followed by the MCP server name and your instructions, e.g., "@VidMCP Segment the person and add a cosmic glow behind them" That's it! The server will respond to your query, and you can continue using it as needed. Here is a step-by-step guide with screenshots.

VidMCP

by dhirajlochib

Overview Schema Source Related Servers Discussions Admin

Python

Local

Creator polish pipeline

One call for a publish-ready talk-head:

orient → denoise → BGM → captions → optional BG replace → export preset

Tool	What it does
`run_talking_head_polish`	Full recipe in one shot
`process_audio`	Denoise · gate · EQ · loudnorm
`mix_bgm`	Ambient pad under voice (ducking)
`transcribe_and_caption`	Whisper words + brand burn-in
`replace_background`	Matte + space / blur / solid plate
`smart_cut_hesitations`	Dead air & filler removal
`export_render`	`youtube_16x9` · `reels_9x16` · `square_1x1`
`import_video`	Bakes portrait rotation upright by default

from vidmcp.tools.creator import run_talking_head_polish

print(run_talking_head_polish(
    "talk.mov",
    preset="reels_9x16",
    bg_mode="space",   # none | space | blur | solid
    strength=0.75,
    bgm_volume=0.35,
))

MCP agents can also call apply_recipe(..., recipe_name="talking_head_polish") or free-text run_intent(...).

Agent context (tool packs + compact results)

Agents drown when MCP hosts load ~90 tool schemas and fat JSON. VidMCP defaults to a small surface:

Setting	Default	Purpose
`VIDMCP_TOOL_PACK`	`talking_head`	Expose only the tools for that product path
`VIDMCP_COMPACT`	`1`	Drop heavy keys / truncate long payloads
`VIDMCP_MAX_RESULT_CHARS`	`4000`	Hard cap on serialized tool results

Packs: talking_head · education · vfx · admin · all

run_intent("polish talk head for reels, space bg, cut fillers", video_path="talk.mov")
project_brief(project_id)          # prefer over full get_project
list_tool_packs() / set_tool_pack("vfx")
get_project(id, detail=true)       # full manifest only when needed

Recipes: talking_head_polish · talking_head_reels · talking_head_space · talking_head_tight · talking_head_infographics.

For full surface (debug): VIDMCP_TOOL_PACK=all.

Related MCP server: video-editor

Gallery

Sample	GIF	MP4
Creator polish	gif	mp4
Flow field	gif	mp4
Tesseract	gif	mp4
Behind the subject	gif	mp4
Kinetic	gif	mp4
Unit circle	gif	mp4
Pipeline	gif	mp4

Install

Needs: Python 3.11+, ffmpeg, uv recommended

uv tool install vidmcp
vidmcp --doctor

# or
pip install 'vidmcp[creator]'   # + faster-whisper + mediapipe

MCP (Grok / Claude / Cursor)

# Grok
grok mcp add vidmcp \
  -e VIDMCP_SAM_BACKEND=mock \
  -e VIDMCP_WORKSPACE_ROOT=$HOME/vidmcp-workspaces \
  -e VIDMCP_TOOL_PACK=talking_head \
  -e VIDMCP_COMPACT=1 \
  -- uvx vidmcp

{
  "mcpServers": {
    "vidmcp": {
      "command": "uvx",
      "args": ["vidmcp"],
      "env": {
        "VIDMCP_SAM_BACKEND": "mock",
        "VIDMCP_WORKSPACE_ROOT": "/path/to/workspaces",
        "VIDMCP_TOOL_PACK": "talking_head",
        "VIDMCP_COMPACT": "1"
      }
    }
  }
}

Optional real matte on Apple Silicon:

pip install 'vidmcp[mlx]'
export VIDMCP_SAM_BACKEND=mlx
export VIDMCP_MLX_MODEL_ID=mlx-community/sam3.1-bf16

Core product


Segment	Text-prompt subject matte (SAM / MLX / MediaPipe / mock)
Compose	Effects behind the subject
Audio	Denoise, enhance, BGM duck, loudnorm
Captions	Whisper timeline + brand burn-in
Educate	Math plates + speech-locked scenes
Export	16:9 · 9:16 · 1:1 without stretch

create_project → import_video (orient)
  → process_audio → mix_bgm
  → segment_subject | replace_background
  → transcribe_and_caption → export_render

Prefer high-level paths: run_talking_head_polish · apply_recipe · run_quality_gated_pipeline.

Project layout

src/vidmcp/     MCP server, creator tools, perception, effects
demos/samples/  GIFs · MP4s for README & site
site/           vidmcp.com (static Hostinger)
configs/        Claude / Cursor snippets

Author

Dhiraj Lochib — Full-Stack, Blockchain, AI
dhirajlochib.com · GitHub · LinkedIn · dhirajch145@gmail.com

License

MIT

Available Tools

27 tools

add_graphicsD

Brand mograph overlays: lower_third, title_card, stat_counter, charts... lock='speech:'.

Inputs: items (required), project_id (required), brand. Returns: object

add_speech_infographicsC

Burn keyword/number infographic cards locked to transcript timing.

Inputs: project_id (required), topic. Returns: object

apply_cut_planC

Execute a cut plan with room-tone fill (no dead digital silence at cuts).

Inputs: plan_id (required), project_id (required), room_tone. Returns: object

apply_recipeC

Run a named production recipe (cyberpunk_talking_head, cinematic_bokeh, product_spotlight, rain_noir, ...).

Inputs: video_path (required), recipe_name (required), project_name, max_render_frames. Returns: object

create_projectC

Create a new non-destructive editing project. Optionally import a source video path.

Inputs: name, video_path. Returns: object

export_edlC

Export edit decision list JSON (source, audio pipeline, renders, history).

Inputs: project_id (required). Returns: object

export_multiB

One call → all platform variants (codec/size/loudness per target, reframe-aware).

Inputs: project_id (required), hw, targets. Returns: object

export_renderC

Export last render (or path) to preset: youtube_16x9 | reels_9x16 | square_1x1 | source.

Inputs: project_id (required), preset, loudnorm, render_path. Returns: object

generate_thumbnailC

Grab mid-frame thumbnail with title overlay from transcript.

Inputs: project_id (required), title. Returns: object

generate_thumbnailsC

A/B thumbnails: face/sharpness/color-scored frames + brand title styles + contact sheet.

Inputs: project_id (required), n, brand, title_variants. Returns: object

get_backend_infoA

Report active SAM backend (MLX/official/ultra/mock), multiplex flags, device, harness.

Inputs: no named parameters. Returns: object

get_projectB

Project info. Default is compact brief; detail=true returns full manifest (large).

Inputs: project_id (required), detail. Returns: object

import_videoA

Import (copy) a video into the project source folder.

bake_orientation: when True (default), bake displaymatrix rotation so portrait phone clips are upright for all downstream tools.

Inputs: project_id (required), video_path (required), bake_orientation. Returns: object

list_projectsA

List all projects in the workspace with status and source paths.

Inputs: no named parameters. Returns: object

list_recipesA

List named production recipes available to apply_recipe.

Inputs: no named parameters. Returns: object

list_tool_packsA

List agent tool packs: talking_head (default), education, vfx, admin, all.

Inputs: no named parameters. Returns: object

mix_bgmC

Mix copyright-safe ambient BGM under vocals (or user bgm_path). Ducking optional.

Inputs: project_id (required), duck, style, volume, +1 more. Returns: object

plan_cutsC

Cut plan v2: retakes, contextual fillers, dead-vs-dramatic pauses. Review before apply.

Inputs: project_id (required), aggressiveness, keep_dramatic_pauses. Returns: object

platform_healthA

Report installed backends (Whisper/SAM/Manim/Blender/FFmpeg) and readiness for education vs GPU paths.

Inputs: no named parameters. Returns: object

process_audioB

Denoise + enhance vocals + loudnorm. Writes audio/vocals_clean.wav on the project.

Inputs: project_id (required), strength, target_lufs. Returns: object

project_briefA

Compact project summary for agents (status, paths, next steps). Prefer over get_project.

Inputs: project_id (required), detail. Returns: object

replace_backgroundC

Cut out subject (MediaPipe/SAM) and composite over plate: space|blur|solid|image.

Inputs: project_id (required), plate, plate_image, matte_backend. Returns: object

run_intentA

Route free-text intent to the right pipeline (polish / recipe / education / VFX).

Examples: 'polish talk head for reels with space bg', 'tight cut no filler', 'cyberpunk behind speaker', 'teach Bayes with math plate'. dry_run=true returns the plan only. video_path required to execute.

Inputs: intent (required), preset, bg_mode, dry_run, +2 more. Returns: object

run_talking_head_polishA

ONE-SHOT creator polish: orient → denoise → BGM → optional cut/BG/captions/cards → export.

Inputs: video_path (required), preset, bg_mode, mix_bgm, +10 more. Returns: object

set_tool_packA

Switch active tool pack and remove out-of-pack tools from this server process.

Expanding to a larger pack (e.g. all) after a smaller pack requires server restart with VIDMCP_TOOL_PACK set — removed tools cannot be re-registered mid-session.

Inputs: pack (required). Returns: object

smart_cut_hesitationsC

Remove dead air / long fillers (um, basically, …) using transcript + energy gaps.

Inputs: project_id (required), aggressiveness. Returns: object

transcribe_and_captionB

ASR word timeline + ASS captions; optionally burn into a new render (brand|karaoke|minimal).

Inputs: project_id (required), burn, style, language, +2 more. Returns: object

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

Deploy Server

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

TDQS

C2.9/5.0

Scored 2026-07-10 23:30 across 27 tools

Disambiguation3/5

Several tools have overlapping purposes, such as plan_cuts, apply_cut_plan, and smart_cut_hesitations all dealing with cutting/trimming, and generate_thumbnail vs generate_thumbnails. Descriptions help somewhat but agents may misselect.

Naming Consistency3/5

Tool names use a mix of verbs (add, apply, create, export, generate, get, import, list, mix, plan, process, project_brief, replace, run, set, smart_cut, transcribe) without a consistent pattern, making it harder to predict functionality.

Tool Count3/5

With 27 tools, the count is high but partially mitigated by the tool pack system that reduces the active set. However, the total number still feels heavy for a single server.

Completeness5/5

The tool set covers the full workflow from project creation to export, including audio, graphics, thumbnails, recipes, and health checks. There are no obvious gaps for its intended domain of video editing automation.

Maintenance

ActivityInactive

ResponsivenessNo recent community issues

Resources

Related MCP Connectors

arsaze-mcpA
io.github.AnrosPrac
AI video editor for agents and humans: timeline, captions, color, audio and generation as MCP tools.
Image & Video Processing Multimedia Processing
251
https://api.arsaze.com/mcp
mcp-wan
io.github.AceDataCloud
MCP server for Wan AI video generation
AI & Machine Learning Image & Video Processing
21
https://wan.mcp.acedata.cloud/mcp
levea-mcp-serverB
io.github.brajendrak00068
Autonomous AI video editor over MCP: viral clips, captions, reframe, chroma key, MP4 export
Image & Video Processing Multimedia Processing Autonomous Agents
15
https://api.livecore.ai/api/v1/misc/mcp
CutClip.oneC
one.cutclip
AI video editor your agent can drive: plan scenes, edit a multi-track timeline, and render MP4s.
App Automation Image & Video Processing Multimedia Processing
1111
https://cutclip.one/api/mcp

Related MCP Servers

media-mcpofficial
Multimedia Processing Image & Video Processing
avclabs
A
license
B
quality
B
maintenance
MCP server for video enhancement and SAM3 image segmentation, enabling tasks like upscaling videos and segmenting objects in images via natural language.
Updated 12 days ago (2026-11-01 15:40 UTC)
4
49 npm
MIT
video-editor
Multimedia Processing Image & Video Processing
followthepattern
A
license
Not graded
quality
D
maintenance
An MCP server for programmatic video editing using ffmpeg, enabling draft creation and refinement via natural language.
Updated 5 months ago (2026-06-20 12:17 UTC)
6 npm
ISC
Svetovid
Multimedia Processing Image & Video Processing
hello-emrah
A
license
A
quality
C
maintenance
An MCP server that lets AI agents visually perceive and edit video inside DaVinci Resolve Studio by rendering frames and manipulating the timeline.
Updated a month ago (2026-10-17 01:05 UTC)
2
1
MIT
video-editor
Multimedia Processing Image & Video Processing
WillBe89
A
license
Not graded
quality
C
maintenance
MCP server for OmniCut that provides 29 deterministic video editing tools (trim, transition, render, etc.) for AI agents to drive the video editor end-to-end.
Updated 3 months ago (2026-08-08 01:18 UTC)
6 npm
1
ISC

Appeared in Searches

영상 (videos/images)

Creator polish pipeline

Agent context (tool packs + compact results)

Gallery

Install

MCP (Grok / Claude / Cursor)

Core product

Project layout

Author

License

Available Tools

Tool Schema Changelog

TDQS

Maintenance

Resources

Related MCP Connectors

Related MCP Servers

media-mcpofficial

video-editor

Svetovid

video-editor

Appeared in Searches