multimedia-processing.md

🖼️ Multimedia Processing

Servers focused on generating or manipulating images, processing video, audio transcription, text-to-speech, or document conversion.

• Pic2SVG: Local stdio MCP server and CLI that uploads raster artwork to Pic2SVG for deterministic SVG tracing. One tool, trace_image_to_svg; filled paths, not centerlines. Install from the source repository with Node.js 20+, then run node /absolute/path/to/pic2svg-mcp/pic2svg-mcp.mjs. No API key; service limits apply. MIT client; separate hosted-service terms. Docs: https://pic2svg.com/mcp/.

• Quady3D: Official local stdio server for text- or single-image-to-3D assets, credit quotes, original-task recovery and saving model files into allowed project folders. Install: npx --yes --ignore-scripts --package=@quady3d/mcp@0.1.2 quady3d-mcp --root /absolute/path/to/project; Node.js 22+, QUADY3D_API_KEY, account credits and download access required. Eight tools. MIT client. Setup guide.

• imageat: MCP server for AI image generation and editing, video, and 20+ image-edit tools. Run with npx -y @imageat/mcp and an IMAGEAT_API_KEY, or use the hosted Streamable HTTP server at https://mcp.imageat.com/mcp (OAuth). Credits per generation. MIT.

• toolkb-mcp: Local stdio server that returns YouTube video and playlist transcripts with timestamps, web pages as Markdown, and App Store reviews with a rating histogram. Install: npx -y github:jack-reson8/toolkb-mcp; Node.js 18+, no API key or account. MIT.

• Snipaya: Local stdio bridge to Snipaya, a macOS screenshot and image annotator. Lets agents add images, draw arrows, numbered steps and callouts, blur secrets, lay out several screenshots on one canvas and export. Runs on the Mac only, no cloud. Install: npx -y snipaya-mcp.

• Doseedo: Official hosted music and audio server for AI agents. doseedo_render_midi turns MIDI to audio, playing the notes an agent writes on 50+ AI instruments (orchestral strings, brass and woodwinds, piano, guitar, drums and World instruments); doseedo_instruments is the open catalog with each instrument's exact call and range; doseedo_generate_sound makes sound effects (SFX) and foley; doseedo_train_instrument makes a private custom instrument from a 30 s to 3 min recording; doseedo_render_score and doseedo_mix render, mix and master a whole score. Plus stem separation (a vocal remover), audio to MIDI transcription, key/BPM/chord detection and DAW session tools (read, edit and download a session as a Logic project; convert projects between ten DAW formats). Hosted Streamable HTTP at https://api.doseedo.com/mcp; OAuth 2.1 with dynamic client registration, or an X-API-Key header. Free tier. Official MCP Registry: com.doseedo/doseedo.

• harshmaur/apify-youtube-transcript-scraper: Gives agents caption text and timestamped segments from YouTube videos, channels and playlists, with optional translation, SRT/VTT and video metadata. Hosted Streamable HTTP at https://mcp.apify.com/?tools=harshmaur/youtube-transcript-scraper; requires an Apify sign-in or token. Charged per successful transcript. MIT.

• Vidova: Official hosted server for Vidova, an agentic video editor. Open projects, read and patch the timeline, cut by transcript, track faces, generate video, images, music, speech and sound effects, check the film and render a finished video. Hosted Streamable HTTP at https://mcp.vidova.ai/mcp; OAuth 2.1 sign-in, no API key. Plugin and skills: https://github.com/youneslaaroussi/vidova-skills (MIT). Official MCP Registry: ai.vidova/vidova.

• Yap MCP: Local stdio server that transcribes audio and video files on your computer with NVIDIA Parakeet TDT v3 (25 European languages, no API key, no upload) and searches your Yap dictation history. Works with Claude Code, Claude Desktop, Cursor, VS Code, Windsurf and Gemini CLI. Install: npx -y github:founden-bot/yap-mcp. MIT.

• Modelflare: Official stdio server for Modelflare, an OpenAI-compatible AI API gateway. List the image and video models your key can call, generate or edit images, start asynchronous video tasks, check their status and download finished videos, with one Modelflare Smart API Key. 5 tools. Install: npx -y @modelflare/mcp-server; auth: MODELFLARE_API_KEY. Docs: https://docs.modelflare.dev. MIT.

• CueFrame: Compose, edit, preview and render videos with captions, graphics and audio. Hosted Streamable HTTP at https://api.cueframe.ai/v1/mcp; OAuth account authorization required. Public plugin, workflow skills and setup documentation: Apache-2.0. The hosted server implementation is separate.

• Lampo: Frame-exact video review for AI agents: read notes and marked screenshots on exact frames, put up new versions, answer on the frame, and wait for the next round until the person approves. Hosted: add the connector https://app.lampo.video/mcp (OAuth sign-in) in Claude, ChatGPT, Cursor or any MCP client. Self-hosted or local (Node ≥22.18, ffmpeg): git clone https://github.com/lampo-vr/lampo.git && cd lampo && npm install, then ./bin/lampo-mcp (stdio). Clients: Claude Code, Codex, Cursor, VS Code, Claude, ChatGPT. Docs: https://github.com/lampo-vr/lampo/blob/main/docs/mcp.md. License: AGPL-3.0-only.

• Fattly: Official hosted server for Fattly, an AI content studio with 50+ models. Generate images, videos, UGC video ads with lip-synced voiceover, voice clones and transcripts; upscale images and remove backgrounds; check credits and job status. Hosted Streamable HTTP at https://fattly.app/api/mcp (OAuth 2.1); local stdio via npx -y fattly mcp with an API key (MIT).

• apimodels.app: Official server for apimodels.app, an OpenAI-compatible API for 150+ image, video, audio and LLM models. Generate or edit images (returns a preview the model can inspect, plus a review_image critique tool), start text-to-video jobs and collect them with get_task, synthesize speech, chat with any listed LLM, list models and check balance. 9 tools. Hosted Streamable HTTP at https://api.apimodels.app/mcp; OAuth sign-in, no key to copy. Local stdio: npx -y apimodels-mcp with APIMODELS_API_KEY. Pay-as-you-go. Docs. Official MCP Registry: io.github.stimQQ/apimodels-mcp. MIT.

• MuAPI: Official hosted MCP server for image, video, and audio generation, model discovery, and asynchronous job results. Streamable HTTP at https://api.muapi.ai/mcp; authenticate with a MuAPI API key as a bearer token.

• Realtime Avatar MCP: Official stdio server for Realtime Avatar, a realtime AI avatar API (live voice and video characters). Lets a coding agent list your avatars, their clip libraries, credit balance and itemised session bill (read-only tools annotated readOnlyHint). Opt-in write tools via REALTIME_AVATAR_ALLOW_WRITES=1 build an avatar from one image, upload assets, declare clip libraries and start a test call (billed per second, refuses production keys). Install: npx -y realtime-avatar-mcp; auth: REALTIME_AVATAR_API_KEY (free sandbox key at https://realtimeavatar.ai/signup). Docs: https://realtimeavatar.ai/docs/mcp. MIT.

• GrowingUpVideo: Turn photos of a person, couple, or pet across the years into an AI growing-up morph video. Agents can get current prices and inclusions with get_offer, provide photo-selection guidance with photo_tips, and return a link to start the project.

• Imagelato: Review image projects, processed batches, deterministic resize and format presets, and webhook delivery health; maintain presets with confirmation. Hosted Streamable HTTP at https://mcp.imagelato.com/mcp; OAuth account authorization required. Docs. Public connector configuration and workflow skills: Apache-2.0.

• thirds.ai: Make branded PDFs and PNG, JPEG, or WebP images from saved templates, brand kits, and data. Hosted Streamable HTTP at https://thirds.ai/mcp; OAuth 2.1 or bearer API key. Official vendor-maintained service.

• FXI Studio: Official hosted server for FXI Studio, a browser-based AI film studio. Generate cinematic images and video clips, frame-chain 2 to 8 shots so each clip starts on the last frame of the one before, compose a scene from saved characters, check credits and look up past generations. 8 tools. Streamable HTTP at https://mcp.fxi.studio/mcp; OAuth 2.1 with dynamic client registration, no API key. Free signup credits, then paid plans. Official MCP Registry: io.github.fueledximagination/fxi-studio.

• Faceless: Official hosted server for Faceless.so. Write a script or pick a topic, generate a finished faceless video with voiceover, visuals, captions and music, then publish or schedule it to YouTube, TikTok, Instagram, X, Facebook, LinkedIn and Threads; can also run automated series and read analytics. Streamable HTTP at https://faceless.so/api/v1/mcp with Authorization: Bearer <FACELESS_API_KEY>, or stdio via npx -y faceless-mcp. Paid plans from $29/month.

• Revid: Official hosted server for Revid. Research what goes viral, make finished short-form videos (voice, captions, music, AI footage) and post them to TikTok, YouTube, Instagram and X. Search viral videos for ideas, render a video from a prompt, script, article or upload, run recurring series, then publish or schedule to connected accounts. 30 tools. Hosted Streamable HTTP at https://www.revid.ai/api/mcp; OAuth 2.1 with dynamic client registration, or a Revid API key as a bearer token. Official MCP Registry: ai.revid/revid.

• Claude Imagine: Hosted remote MCP server (Streamable HTTP at https://claudeimagine.com/api/mcp, OAuth 2.1 with dynamic client registration, no API key) for generating and editing images and videos with Nano Banana 2, GPT Image 2.5, Seedream 4.5, Flux 2 Pro and Veo 3.1 Fast; quote_generation returns the credit cost before a call. Free signup credits, then paid plans. Independent service, not affiliated with Anthropic. Docs: claudeimagine.com/docs/mcp.

• SHAR Production Metadata MCP: Read-only stdio MCP server for AI-hybrid video production manifest validation. validate_production_manifest accepts a production delivery manifest and returns structural validation plus a releasability result without reading local files, calling networks, or publishing data. Install from source with git clone https://github.com/SHARProduction/production-metadata-mcp.git && cd production-metadata-mcp && npm ci, then run node server.js. No auth; MIT.

• ChevalGrand520/local-gpu-imagegen: Auditable, confirmation-gated local GPU image generation for existing ComfyUI installations, with cryptographic SHA-256 model identity, explicit license approval, no silent model downloads, and durable per-run evidence. Supports Windows 10/11 x64 with NVIDIA GPUs. 17 tools. stdio via uvx --from local-gpu-imagegen==0.9.1 local-gpu-imagegen serve; no authentication. MIT.

• pratie/bulktranscripts-mcp: Hosted YouTube server for bulk transcript extraction plus search, channel and playlist listings, and free new-upload tracking via RSS. Accepts video URLs or 11-character ids, returns plain text or timestamped segments, and batches up to 20 videos per call; already-fetched transcripts re-read free from the caller's library. 7 tools. Streamable HTTP at https://bulktranscripts.co/mcp; keyless on the free tier (30 transcripts per IP), optional Authorization: Bearer <license key> for one-time credit packs. Free tier. MIT.

• haljishi/vidwords-mcp: Hosted YouTube server that searches one video's transcript and returns only the matching moments with clickable ?t= deep links, plus frame-level analysis of slides, charts, demos and on-screen text; citations are verified against a recorded frame or a real transcript segment and dropped if they fail. 9 tools. Streamable HTTP at https://vidwords.com/mcp; Authorization: Basic <token> header, or OAuth 2.1 for claude.ai and ChatGPT. Free tier. MIT.

• YouTube Transcript Scraper (changefeeds, via Apify MCP): Hosted YouTube transcript tool. Accepts video, Shorts, channel or playlist URLs (or bare ids) and returns manual or auto-generated captions as timestamped segments and/or plain text, with title, channel, duration, views and publish date. Requests go through Apify residential proxy, so it works from cloud hosts where YouTube's bot check blocks direct fetches. One Actor tool, changefeeds--youtube-transcript-scraper, plus Apify's run and dataset helpers. Streamable HTTP at https://mcp.apify.com/?tools=changefeeds/youtube-transcript-scraper; auth: Apify OAuth or Authorization: Bearer <Apify API token>. Paid: $0.004 per transcript returned, billed by Apify (videos without captions are not charged). Official MCP Registry id: io.github.Skyblueballykid/youtube-transcript. Source is private; license unknown.

• magichourhq/magic-hour-mcp: Official hosted Magic Hour server for generating and editing images, videos, and audio, including source-media uploads, asynchronous job polling, and signed download URLs. 44 tools. Streamable HTTP at https://mcp.magichour.ai/; Authorization: Bearer <MAGIC_HOUR_API_KEY>.

• PSPDFKit/nutrient-dws-mcp-server: Nutrient Document Web Services MCP server for PDF and Office document workflows including conversion, merge, redaction, signing, OCR, watermarking, and extraction.

• smartdatabrokers/slideforge-mcp: Hosted MCP server for consulting-grade PowerPoint generation from typed slide intents or a prose brief. Renders fully editable .pptx decks — charts, diagrams, brand templates — deterministically, with a machine-readable fidelity manifest on every response. 7 tools. Streamable HTTP at https://api.slideforge.dev/mcp/; OAuth 2.1 for Claude Desktop and bearer/API-key auth for other clients. MIT.

• MohamedAbdallah-14/prompt-to-asset: Generates production-ready visual assets (app icons, favicons, OG images, logos, wordmarks) by routing each request across 30+ image generation models to the best fit. Zero-key first run via free-tier providers (Pollinations, Stable Horde, HuggingFace). MIT licensed.

• AIDC-AI/Pixelle-MCP: An omnimodal AIGC framework that seamlessly converts ComfyUI workflows into MCP tools with zero code, enabling full-modal support for Text, Image, Sound, and Video generation with Chainlit-based web interface.

• SkyworkAI/Mureka-mcp: Facilitates the creation of lyrics, songs, and background music through an MCP server, enabling seamless integration with platforms like Claude Desktop and OpenAI Agents.

• omergocmen/json2video-mcp-server: Facilitates video creation and status monitoring through the json2video API, enabling seamless integration with LLMs and automation agents.

• echozyr2001/ali-flux-mcp: Facilitates image generation and management using Alibaba Cloud's DashScope API, with task tracking and local storage capabilities.

• c-rick/jimeng-mcp: A TypeScript-based MCP server integrating Volcengine's AI image generation service, offering tools for creating images with customizable parameters and direct URL returns.

• hamflx/imagen3-mcp: Harnesses Google's Imagen 3.0 for photorealistic image generation via MCP, requiring a Google Gemini API key.

• KyaniteLabs/mcp-video: AI-native video editing MCP server for trimming, merging, text overlays, audio, subtitles, effects, quality checks, and FFmpeg-backed workflows.

• OrkasVideoStudio: Local-first TypeScript MCP and CLI toolkit for agent-driven video composition, editing, generation, transcription, and assembly from editable plan.json timelines. Install from source while its npm packages are being published.

• PromoShot: Open Rust promo-video engine as MCP — author .promo layer/keyframe projects (screen recordings, photos, captions), validate, and render stills/GIF/mp4 via wgpu + ffmpeg. Same compositor as the Mac/iOS apps. Install: releases or docker pull ghcr.io/garalex/promoshot-mcp. Transport: stdio. Docs: https://promoshot.app/mcp. License: Apache-2.0.

• SealinGp/mcp-video-extraction: Facilitates text extraction from videos and audio files across multiple platforms using OpenAI's Whisper model.

• kdr/mcp-draw: Facilitates AI-driven image generation from text prompts via a standardized interface.

• 4kk11/mcp-gpt-image: Generates and edits images using OpenAI API, providing scalable previews and Docker integration.

• antvis/mcp-server-chart: Facilitates the creation of diverse visual charts using AntV through a TypeScript-based MCP server.

• HYPERVAPOR/mcp-image-processor: High-performance image processing server offering format conversion, resizing, and optimization capabilities.

• Bigchx/mcp_3d_relief: Transform 2D images into detailed 3D relief models in STL format for 3D printing or rendering.

• falahgs/mcp-3d-style-cartoon-gen-server: A server that combines 3D-style cartoon image generation with secure file system operations, leveraging Google's Gemini AI and MCP SDK.

• zjf2671/hh-mcp-comfyui: Facilitates image generation through natural language commands by interfacing with a local ComfyUI instance via the MCP protocol.

• mario-andreschak/mcp_video_recognition: Facilitates image, audio, and video recognition using Google's Gemini AI.

• bads1de/youtube-mp3-mcp: Facilitates the extraction of high-quality MP3 audio from YouTube URLs with seamless Claude Desktop integration.

• jyjune/mcp_vms: Facilitates integration with CCTV systems by providing tools to access and control video streams and PTZ cameras.

• nguyendinhsinh361/elevenlabs-mcp: Facilitates interaction with ElevenLabs' Text to Speech and audio processing APIs, enabling MCP clients to generate speech, clone voices, and transcribe audio.

• tjh19971228/mcp_video_analysis: Facilitates video content analysis and mind map generation using the Model Context Protocol.

• intsig-textin/textin-mcp: TextIn MCP Server facilitates text extraction and OCR on documents, supporting recognition and conversion to Markdown format.

• PixVerseAI/PixVerse-MCP: Enables seamless interaction with PixVerse's AI video generation models through applications supporting the Model Context Protocol.

• falahgs/mcp-minimax-music-server: Facilitates AI-driven music and audio content creation through the MiniMax Music API, seamlessly integrating with Claude Desktop.

• morim3/mcp_adobe_premiere: Facilitates LLM control over Adobe Premiere Pro via an MCP server and UXP plugin integration.

• savethepolarbears/google-photos-mcp: Facilitates AI assistants' access to Google Photos, enabling photo search and retrieval by content, date, and location.

• luebken/playlist-mcp: Provides access to YouTube playlist transcripts via an experimental MCP server, with initial support for KubeCon London 2025 content.

• sandst1/mcp-server-midi: Facilitates the transmission of MIDI sequences from an LLM to any MIDI-compatible software, enabling seamless integration with digital audio workstations and hardware synthesizers.

• falahgs/image-gen3-google-mcp-server: Harness Google's Imagen 3.0 model via the Gemini API for high-quality image generation, seamlessly integrating with Claude Desktop and other MCP-compatible hosts.

• mang0cola/watermark-mcp-server: Enhance images by adding customizable text or image watermarks using a simple MCP server.

• falahgs/flux-imagegen-mcp-server: A specialized server for generating and manipulating images using Pollinations AI, compliant with the Model Context Protocol.

• nikmaniatis/Pd-MCP-Server: Facilitates dynamic interaction between Claude AI and Pure Data for real-time music creation through natural language commands.

• MichaelYangjson/mcp-ghibli-video: A TypeScript-based server offering AI-driven image and video generation with Ghibli-style animations.

• undertaker86001/mcp-process-pdf: A robust MCP server for processing PDF documents with features like text extraction, image optimization, and intelligent classification using deep learning.

• Synohara/supercollider-mcp: Facilitates the execution of SuperCollider synths using supercolliderjs through an MCP server.

• slot181/sd-image-gen-mcp: Provides text-to-image generation using Stable Diffusion WebUI API with optional Cloudflare ImgBed integration.

• coderjun/shaka-packager-mcp-server: Integrates Shaka Packager with Claude AI for video transcoding, packaging, and analysis, enhancing media processing capabilities.

• rjn32s/mcp-ocr: A robust OCR server leveraging MCP to extract text from images using Tesseract, supporting various input types and languages.

• An-3/mcp-audacity: Facilitates remote control of Audacity through MCP endpoints using named pipes and the mod-script-pipe interface.

• Anna-Pinewood/mcp-pallete: Generates color palettes from images using an MCP server with IMAGGA integration.

• shuntagami/elevenlabs-scribe-transcriber-ts: A TypeScript tool for transcribing audio and video files using the ElevenLabs Scribe model, with MCP server capabilities for integration with clients like Claude Desktop.

• mcai/podcast-tts-mcp: Generate multilingual podcast conversations with alternating male and female voices using Microsoft Edge's TTS technology.

• HANON-games/midi-control-mcp: Facilitates MIDI message transmission to output devices via a TypeScript-based MCP server.

• grentank/mcp-server-gigachat-image-generation: Facilitates image generation through a Dockerized MCP server with Gigachat integration.

• mjpstj12/abletonOsc_mcp: Facilitates communication between LLMs and Ableton Live using OSC for music production and session manipulation.

• xhiroga/blender-mcp-senpai: Facilitates integration with Blender by indexing documents, searching with DuckDB-VSS, and utilizing RAG through MCP.

• andyanalog/mcp-ffmpeg-livestream-aws: A toolkit for FFmpeg video processing and live streaming with AWS integration.

• 302ai/302_image_mcp: Facilitates image processing tasks through an MCP server, integrating seamlessly with Claude Desktop.

• karoterra/aviutl-mcp: Facilitates the control of AviUtl through MCP, enabling seamless integration with various MCP hosts like Claude Desktop.

• Yoshino-Yukitaro/imagemagick_mcp_server: Facilitates image processing tasks using ImageMagick through an MCP server setup.

• kachiO/mlx-whisper-mcp: Facilitates audio transcription and YouTube video transcription using MLX Whisper on Apple Silicon Macs.

• instructa/mcp-youtube-music: Facilitates searching and playing YouTube Music tracks through AI assistants like Cursor or Claude Desktop.

• shivammaurya042/mcp-render: Effortlessly manage your Render.com account with this MCP server, enabling seamless deployment management and environment variable handling through your preferred MCP client.

• hanqizheng/media_kit_mcp_server: Facilitates the parsing of Media Kit PDF files for key content extraction in the advertising and marketing sector.

• djbriane/plex-mcp: Integrates with Plex Media Server API to manage movies and playlists using a Python-based MCP server.

• GoatWang/YTTranscipterMultilingualMCP: Transcribe YouTube videos into multiple languages using a multilingual MCP server.

• tubone24/midi-mcp-server: Enables AI models to generate MIDI files from text-based music data, facilitating programmatic musical composition through a standardized interface.

• grafikogr/freepik-mcp-server: Generates images from text descriptions using Freepik's Flux AI service, integrated with Claude Desktop.

• diivi/aseprite-mcp: Facilitates interaction with the Aseprite API through an MCP server, enabling automated graphic manipulations.

• Henry-The-Yang-101/spotify_mcp: Connects to the Spotify Web API to provide playback and search functionalities to AI clients via MCP.

• palvindersander/youtube-transcript-mcp: Fetches YouTube video transcripts and metadata for seamless integration with Claude, including fact-checking capabilities.

• minbang930/Youtube-Vision-MCP: Utilizes the Google Gemini Vision API to analyze and interact with YouTube videos, offering features like summarization, Q&A, and key moment extraction.

• fl0w1nd/grok2-image-mcp-server: Facilitates image generation using the Grok-2 model via the MCP protocol for chat assistants.

• x007xyz/image-process-mcp-server: Facilitates image manipulation tasks such as resizing, format conversion, cropping, and rotation using the Sharp library.

• demon24ru/video-transcribe-mcp: Integrates with optivus to provide video transcription capabilities for LLMs, supporting platforms like YouTube, Facebook, and Tiktok.

• demon24ru/fish-speech-mcp: Facilitates text-to-speech synthesis for LLMs with FishSpeech integration.

• R-lz/mcp-video-digest: Facilitates video transcription and summarization from platforms like YouTube and Bilibili using multiple transcription services.

• longbowzz/svg2png_mcp: Facilitates SVG to PNG conversion using CairoSVG or Inkscape, integrated with MCP protocol for seamless client interaction.

• richbai90/spotify-mcp: Integrates with Spotify API to manage and create playlists through Claude.

• wheattoast11/mcp-video-gen: Facilitates video and image generation using RunwayML and Luma AI APIs, with capabilities to enhance prompts via OpenRouter LLMs.

• IA-Entertainment-git-organization/youtube-video-summarize: Summarizes YouTube videos by extracting and analyzing transcripts using various algorithms, including OpenAI's API for high-quality summaries.

• nabid-pf/youtube-video-summarizer-mcp: Enables Claude to fetch and summarize YouTube videos by extracting titles, descriptions, and transcripts.

• jlevy/deep-transcribe: Facilitates deep transcription of video and audio content with speaker identification and annotations, operable as an MCP server.

• TomokiIshimine/svg-render-mcp: Transforms SVG strings into PNG images with customizable dimensions and background colors.

• wanga90/blender-mcp: BlenderMCP enables Claude AI to interact with Blender for enhanced 3D modeling and scene manipulation through the Model Context Protocol.

• vidau-ai/asr_mcp_server: Provides ASR capabilities using the whisper engine and exposes TTS functionality for seamless speech synthesis integration.

• abhiemj/manim-mcp-server: Facilitates the execution of Manim animation scripts and returns the generated video through an MCP server interface.

• tkoba1974/mcp-kroki: Transforms various diagram formats into images using Kroki.io, supporting formats like Mermaid and PlantUML.

• Nazruden/mcp-openvision: Leverage OpenRouter vision models to enable AI-driven image analysis within the MCP ecosystem.

• thuhoai27/mcp-image-reader: ImageReader converts local image files to Base64-encoded data for AI model analysis.

• 13rac1/videocapture-mcp: Enables AI assistants to control webcams and capture images using OpenCV.

• infinitimeless/radiofrance-podcast-explorer-mcp: Facilitates AI-driven exploration and access to Radio France's podcasts and audio content through a Model Control Protocol server.

• frankdeno/flux-image-generator-mcp: Generate images from text prompts using the FLUX model with customizable settings and batch processing capabilities.

• GaoCan702/mcp-gererate-image: Facilitates AI-driven image generation requests from Cursor using Cloudflare's Flux AI model, with local or temporary storage options.

• video-creator/ffmpeg-mcp: Facilitates local video search, editing, and playback using FFmpeg commands through an MCP server interface.

• sanxfxteam/gemini-mcp-server: Facilitates image generation using Google's Gemini 2 API through an MCP server interface.

• champierre/image-mcp-server: Analyzes image content from URLs or local paths using the GPT-4o-mini model, providing high-precision recognition and description.

• Ichigo3766/image-gen-mcp: Facilitates text-to-image generation using Stable Diffusion WebUI API, enabling creative visual outputs from textual prompts.

• Ichigo3766/audio-transcriber-mcp: Facilitates audio transcription using OpenAI's Whisper API, enabling seamless integration with MCP environments.

• rakeshgangwar/tmdb-mcp-server: Facilitates AI-driven movie searches and information retrieval via The Movie Database API using the MCP interface.

• rocksun/media-mcp-server: Facilitates multimedia metadata management through a RESTful API using the FastMCP framework.

• yasar38/BLENDER-MCP-CURSOR-: Facilitates AI-driven 3D modeling in Blender through seamless integration with Cursor using the Model Context Protocol.

• sinco-lab/mcp-youtube-transcript: Facilitates the extraction and processing of YouTube video transcripts for content analysis.

• blacktop/mcp-say: Facilitates text-to-speech capabilities using macOS 'say' command and ElevenLabs API for integration with MCP protocol-based applications.

• tijs/py-sound-mcp: Enhances coding environments with audio feedback for events using the Model Context Protocol, compatible with Cursor and other IDEs.

• Johnnyhtw/audio2txt_mcp: Transforms meeting audio recordings into text using Whisper, integrating with CLINE for AI-generated summaries.

• abhishekjairath/sonic-pi-mcp: Facilitates AI-driven musical creation by enabling AI assistants to control Sonic Pi through OSC messages.

• crazyrabbitLTC/mcp-vibecoder: Facilitates structured LLM-based coding workflows with feature clarification, task tracking, and document management.

• ejfox/vulpes-spotify-mcp: Facilitates AI assistants in interacting with Spotify for track search and playback.

• surferdot/mcp-svg-converter: Transforms SVG code into high-quality PNG and JPG images with customizable options, ensuring secure file handling and integration with Claude Desktop.

• vinayak-mehta/mcp-sonic-pi: Connects MCP clients with Sonic Pi to create music using English commands.

• IA-Programming/mcp-images: Enterprise-grade image processing server offering tools for fetching and processing images from various sources, ideal for AI applications and data pipelines.

• royshil/obs-mcp: Control OBS Studio through an MCP server using the OBS WebSocket protocol for scene management, source control, and streaming operations.

• yoavniran/cloudinary-mcp-server: Facilitates AI-driven interactions with Cloudinary by exposing its Upload & Admin API methods as callable tools.

• bjkemp/mcp-midi: Facilitates MIDI device interaction through natural language commands, enabling control over synthesizers, note playback, and instrument changes.

• BigUncle/Fast-Whisper-MCP-Server: High-performance speech recognition server leveraging Faster Whisper for efficient audio transcription.

• flipsai/davinci-resolve-mcp: Facilitates AI-driven video editing and color grading in DaVinci Resolve through Claude AI integration using the Model Context Protocol.

• hammeiam/koroko-speech-mcp: Provides high-quality text-to-speech capabilities using the Kokoro TTS model with customizable voice and speed options.

• JavaProgrammerLB/unsplash-mcp-server: Facilitates image searches on Unsplash using a Java-based MCP server.

• Tooflex/davinci-resolve-mcp: Facilitates AI-driven control over DaVinci Resolve Studio for advanced editing and media management.

• obre10off/spotify-mcp: Control Spotify playback using natural language commands through an MCP client.

• ahujasid/ableton-mcp: Connects Ableton Live to Claude AI for interactive music production and session control via the Model Context Protocol.

• marcelmarais/spotify-mcp-server: A lightweight server enabling AI assistants to control Spotify playback and manage playlists.

• rocksun/gemini-image-mcp-server: Provides AI-driven image generation and editing services using the MCP protocol with Google Gemini technology.

• okooo5km/unsplash-mcp-server-swift: Swift-based server enabling LLMs to search and retrieve photos from Unsplash with advanced filtering options.

• samuelgursky/davinci-resolve-mcp: Connects AI coding assistants to DaVinci Resolve, enabling natural language control and querying.

• Rupeebw/mcp-image-reader: A TypeScript-based MCP server that facilitates the creation, management, and summarization of text notes using URIs and metadata.

• Garblesnarff/gemini-mcp-server: Facilitates image generation on Claude Desktop using Google's Gemini AI models.

• eetumartola/houdini-mcp: Facilitates seamless interaction between Claude AI and Houdini for enhanced 3D modeling and scene manipulation.

• elliotxx/favicon-mcp-server: Transforms SVG images into ICO and PNG favicon formats for web applications, supporting seamless integration with LLM-powered tools.

• Ut13158/blender-mcp: BlenderMCP enables seamless interaction between Blender and Claude AI for enhanced 3D modeling and scene manipulation through the Model Context Protocol.

• hugohow/mcp-music-analysis: Leverage librosa and Whisper to analyze music audio through an MCP server, integrating seamlessly with LLMs for enhanced audio insights.

• xue160709/yt-mcp-server: A Model Context Protocol server utilizing the mcp-framework for tool development and integration with Claude Desktop.

• MCERQUA/freepik-mcp: Facilitates interaction with Freepik's API for accessing stock photos and generating images using Mystic AI.

• zaptrem/music-mcp: Facilitates music generation through text prompts using Sonauto's API, integrated with Claude Desktop.

• SIAM-TheLegend/sound-mcp: A lightweight MCP server that remotely triggers customizable notification sounds via AI agents using npx.

• androidStern/ableton-vibe: Facilitates MIDI track creation in Ableton Live through MCP server integration.

• GMKR/mcp-imagegen: Facilitates image generation using Together AI's models via a configurable MCP server.

• bitscorp-mcp/mcp-ffmpeg: A Node.js server leveraging FFmpeg to provide APIs for video resizing and audio extraction, seamlessly integrating with Claude Desktop for natural language video processing.

• everaldo/mcp-mistral-ocr: Facilitates OCR processing of images and PDFs using Mistral AI's API, with support for local and URL-based files.

• GongRzhe/Audio-MCP-Server: Facilitates audio input/output for AI assistants, enabling interaction with computer audio systems for recording and playback.

• VxASI/blender-mcp-vxai: Blender MCP VXAI enables seamless natural language control over Blender for creating and manipulating 3D models and animations, integrating AI-driven automation for enhanced creative workflows.

• CLOUDWERX-DEV/DiffuGen: DiffuGen offers a seamless interface for generating images using Stable Diffusion models, integrating directly with IDEs via MCP for efficient development workflows.

• tomcat65/flux-images-mcp: A server for generating, modifying, and inpainting images using Replicate's Flux models, with features for obtaining image URLs and browsing stored images.

• MushroomFleet/TranscriptionTools-MCP: Enhances transcript processing with intelligent formatting, error repair, and summarization using advanced LLMs.

• zym9863/elevenlabs-sound-effect-server: Generates sound effects from text descriptions using ElevenLabs API, saving them as MP3 files.

• AngeloGiacco/elevenlabs-mcp: Exposes OpenAPI-defined ElevenLabs endpoints as MCP resources for seamless integration with Claude Desktop.

• Krekun/mcp-voicevox: Facilitates text-to-speech conversion using VoiceVox through an MCP server, enabling Claude to generate diverse audio outputs.

• Handwriting-OCR/handwriting-ocr-mcp-server: Facilitates integration between MCP clients and the Handwriting OCR service for document transcription and retrieval.

• haltakov/meme-mcp: Generate memes from user prompts using the ImgFlip API with this MCP server.

• okooo5km/zipic-mcp-server: Swift-based server enabling AI assistants to compress and optimize images using advanced tools and batch processing capabilities.

• bendusy/pollinations-mcp: Connects AI models to Pollinations.ai for image and text generation via MCP protocol.

• gregkop/sketchfab-mcp-server: Facilitates seamless interaction with Sketchfab's 3D model platform, enabling search, detailed viewing, and downloading of models.

• peng-shawn/mermaid-mcp-server: Transforms Mermaid diagram code into PNG images using Puppeteer for seamless AI integration.

• rmcendarfer2017/MCP-image-gen: Connects to Replicate's API to generate and manage images with customizable prompts and styles.

• giannisanni/kokoro-tts-mcp: Facilitates text-to-speech synthesis using the Kokoro TTS engine, enabling seamless integration of speech capabilities into applications via MCP tools.

• supercurses/powerpoint: Facilitates the creation and editing of PowerPoint presentations with dynamic slide generation and multimedia integration.

• Tomocrystal/generate_image-mcp-server: A server implementation for generating images using Model Context Protocol, compatible with OpenAI DALL-E and other large language model APIs.

• rhish9h/image-filter-mcp: Provides 10 image filters using Python and Pillow, integrating seamlessly with Claude Desktop.

• setkyar/youtube-subtitles-mcp: Facilitates AI assistants in downloading and analyzing YouTube video subtitles with seamless integration capabilities.

• DaInfernalCoder/unsplash-server: Integrates Unsplash image search and retrieval with AI assistants, enabling seamless access to high-quality images.

• DeepSRT/deepsrt-mcp: Facilitates YouTube video summarization through DeepSRT API integration, offering narrative and bullet-point summaries with multi-language support.

• manascb1344/together-mcp-server: Enables high-quality image generation using the Flux.1 Schnell model via Together AI.

• Simon-Kansara/ableton-live-mcp-server: Facilitates communication between LLMs and Ableton Live using OSC for seamless music production control.

• hirokidaichi/imark: A versatile CLI tool that leverages AI for image recognition, cataloging, and generation, with MCP server capabilities for enhanced image processing.

• ckz/flux-schnell-mcp: Generates images using the Flux Schnell model via Replicate API with customizable prompts.

• index01d/ytrnscrpt-mcp-server: Facilitates Claude's ability to retrieve and analyze YouTube video transcripts through an MCP server.

• jkawamoto/mcp-florence2: Processes images and PDFs to extract text or generate descriptive captions using Florence-2.

• glifxyz/glif-mcp-server: Official hosted media-generation agent from Glif. Generate images, video, and audio, transcribe media, and chain multi-step creative workflows from any MCP client. Streamable HTTP endpoint: https://glif.app/api/mcp; OAuth 2.1. MIT.

• maoxiaoke/mcp-media-processor: Node.js server for advanced video and image processing using MCP, featuring conversion, compression, and editing capabilities.

• lalanikarim/comfy-mcp-server: Utilizes the FastMCP framework to generate images from prompts by interfacing with a remote Comfy server.

• husniadil/mcp-image-placeholder: Generates placeholder images from multiple providers, offering seamless integration with tools like Claude for Desktop and Cursor.

• dandeliongold/mcp-decent-sampler-drums: Generates DecentSampler drum kit configurations with WAV analysis and XML generation tools.

• kennethreitz/mcp-applemusic: Control Apple Music on macOS using AppleScript commands via a FastMCP server.

• jkawamoto/mcp-youtube-transcript: Retrieve transcripts from YouTube videos using a Model Context Protocol server.

• mondweep/youtube-music-mcp-server: Facilitates AI-driven YouTube Music playback through Chrome, enabling song search and play by song or artist name.

• superseoworld/mcp-spotify: ArtistLens provides seamless access to Spotify's music catalog, enabling advanced search and retrieval of artist and album information through a robust MCP server.

• catalystneuro/mcp_read_images: Facilitates image analysis using OpenRouter vision models with configurable model selection and automatic image optimization.

• felores/placid-mcp-server: Integrate with Placid.app's API to dynamically generate images and videos using customizable templates.

• Here-and-Tomorrow-LLC/audio-player-mcp: Enables Claude to manage audio playback on your computer through a Model Context Protocol server.

• felores/cloudinary-mcp-server: Facilitates media uploads to Cloudinary via Claude Desktop and compatible MCP clients.

• truaxki/mcp-Pdf2png: Facilitates the conversion of PDF documents into PNG images using a Model Context Protocol server.

• Kush36Agrawal/Video_Editor_MCP: A robust video editing server utilizing FFmpeg for executing video operations via natural language commands.

• tadasant/mcp-server-stability-ai: Integrates MCP Clients with Stability AI for advanced image manipulation, including generation, editing, and upscaling.

• apinetwork/piapi-mcp-server: A TypeScript-based MCP server that enables media content generation through PiAPI's API, integrating seamlessly with platforms like Midjourney and Claude.

• zxkane/mcp-server-amazon-bedrock: Integrates Amazon Bedrock's Nova Canvas model for generating high-quality images from text descriptions with advanced control options.

• sparfenyuk/mcp-youtube: Facilitates AI assistants' interaction with YouTube by bridging the YouTube API and AI tools for tasks like downloading captions and summarizing videos.

• spencerhhubert/illustrator-mcp-server: Facilitates script execution on Adobe Illustrator via an MCP server, leveraging JavaScript and AppleScript for MacOS compatibility.

• burningion/video-editing-mcp: Facilitates video editing and management through an MCP interface, integrating with Video Jungle for uploading, editing, and searching videos using LLMs.

• angheljf/dalle-image-server: A TypeScript-based server that generates images using DALL·E 2 from text prompts, requiring an OpenAI API key.

• sammyl720/image-generator-mcp-server: Generates images from prompts using OpenAI's DALL-E-3 model, with TypeScript implementation for seamless integration with Claude Desktop.

• kimtaeyoon83/mcp-server-youtube-transcript: Facilitates direct transcript retrieval from YouTube videos with language-specific options.

• iterationlayer/mcp: Document extraction, image transformation, programmatic image generation, document generation, and sheet generation via composable APIs with OAuth 2.1 authentication.

• neugence/whipscribe-mcp: MCP server for audio/video transcription via the Whipscribe API. Transcribes from a URL (YouTube, podcasts, direct media links) or local file; returns txt/json/srt/vtt/docx with optional speaker diarization and word timestamps. Hosted remote endpoint at https://whipscribe.com/mcp (Bearer auth) or local install via uvx whipscribe-mcp.

• daedalus/mcp-stl - MCP server for editing STL 3D model files.

• rendobar/mcp: Official local stdio MCP server for Rendobar's media-processing API. Lets AI agents submit FFmpeg, transcode, caption, and watermark jobs and upload local files to the cloud backend. TypeScript, MIT licensed.

• getcutpro/mcp: MCP server that exposes the full CutPro v1 API as 34 tools for AI clients — analyze videos, submit clipping jobs, manage clips, render, publish posts, and more. Supports stdio (local), Streamable HTTP (remote), and OAuth 2.1 (browser clients).

• transloadit/node-sdk: Agent-native media processing via Transloadit's 86+ Robots, supporting video encoding, image manipulation, document conversion, OCR, and speech transcription. Hosted or self-hosted via npx.

• fasuizu-br/speech-ai-examples: Speech AI MCP server with pronunciation assessment (phoneme-level scoring, 17MB model, <300ms), text-to-speech, and speech-to-text. 8 tools for AI agents building language learning, accessibility, and voice applications.

• ssembleinc/ssemble-mcp-server: Create AI-generated short-form video clips from YouTube with captions, music, gameplay overlays, meme hooks, and viral scoring.

• pictify-io/mcp: Render images, animated GIFs, and PDFs from HTML/CSS, live URLs, or reusable Handlebars/FabricJS templates. Batch up to 100 personalised renders per request and A/B test image variants with built-in experiments. Hosted remote at https://mcp.pictify.io or self-host via npx -y @pictify/mcp-server.

• demo112/mcp-server-convert: Converts PDF, DOCX, HTML, EPUB, CSV, Excel, PowerPoint, JSON, XML to Markdown via MCP tools with batch processing and metadata extraction.

• AnveVoice: Voice AI agent for websites with agentic DOM actions, MCP support, 50+ languages, <700ms latency, and one-line embed. Free tier available.

• alichherawalla/video-overlay-kit: Render 4-6s animated b-roll overlay videos for short-form social (LinkedIn, IG Reels, YouTube Shorts, TikTok) and landscape video. MCP-driven: paste a script, the AI writes the scene spec and renders the mp4. Free, MIT, runs locally.

• hushvert/mcp: File conversion for AI agents. Office docs to PDF, PDF to Word, document interchange (Markdown/HTML/EPUB/LaTeX/RST), and audio/video transcodes via the hushvert hosted API. 4 tools: convert_file, convert_poll, list_formats, check_usage. Install: npx -y @hushvert/mcp. Stdio transport; API key via HUSHVERT_API_KEY. MIT.

• SylphxAI/anymd: Converts any file to clean Markdown for agents: PDF, DOCX, PPTX, XLSX/CSV, EPUB, HTML/URLs, images (OCR) and audio/video metadata, with read, search and inspect tools. Local Rust binary, no API key. Formerly pdf-reader-mcp. Install: npx -y @sylphx/anymd.

• Pepys (pepys-mcp): Transcribes audio, video, and podcast feeds into speaker-labeled, timestamped transcripts with SRT/VTT output, URL ingestion, batch processing, and transcript search across 99+ languages. Install: npx -y pepys-mcp with PEPYS_API_KEY, or connect to the hosted OAuth endpoint at https://pepys.co/api/mcp. MIT.

• ilovevideoeditor/mcp-server: Render videos from JSON via the iLoveVideoEditor cloud API — motion-design templates, GPU effects, transitions, batch rendering, webhooks. Install: npx -y @ilovevideoeditor/mcp-server (stdio; API key via VF_API_KEY).

• runcomfy-com/runcomfy-mcp: Official remote MCP server for RunComfy. Its 31 tools cover ComfyUI Serverless GPU deployments and asynchronous image/video inference, hosted model inference (browse, run, poll, cancel), and LoRA training jobs (dataset upload, submit/poll/resume/edit training). Connect to https://mcp.runcomfy.com/mcp with a RunComfy API token in the Authorization: Bearer header. MIT.

• BizMuse AI Music Video: Official stdio MCP server for asynchronous AI music video generation. Agents can discover supported models, upload local audio and reference images, submit directed music-video jobs, monitor progress, retrieve outputs, and inspect account credits. Install: npx -y bizmuse-mcp with BIZMUSE_API_KEY. MIT.

• JOYLINK-LTD/lacuna-toolkit: Official MCP server for the Lacuna Music API, enabling AI music generation - full songs with vocals and lyrics or instrumentals - with style prompts, task status polling, and audio downloads.

• mutonby/openshorts: Open-source AI video platform with a built-in MCP server. Turns long videos into viral 9:16 shorts (AI moment detection, face-tracking reframe, animated subtitles, dubbing) and generates UGC videos with AI actors. 6 tools: process_video, get_job_status, list_clips, get_quota, add_subtitles, publish_clip. Streamable HTTP; self-host with Docker at http://localhost:8000/mcp (no key) or use the hosted endpoint https://mcp.openshorts.app/mcp with a Bearer osk_... key. Python, MIT.

• Cadence Code: Fully local voice conversations for coding agents on Apple Silicon. A stdio MCP server owns the selected text-to-speech and speech-to-text models directly through MLX Audio, so speech never leaves the machine and no cloud speech service is used. Models are user-selectable. 8 tools: voice_start, voice_speak, voice_listen, voice_interrupt, voice_stop, voice_models, voice_configure, voice_status. Works with Claude Code, Codex, Cursor, and Antigravity. Install: /plugin marketplace add michael-L-i/cadence-code. Python, MIT.

• humanpen/humanpen-mcp: Document-level AI humanizer for .docx/.pptx — rewrite a whole file, only selected passages, or just the text flagged by a Turnitin/iThenticate AI-detection report, editing in place while preserving formatting, tables, images, citations, and formulas. Install: npx -y humanpen-mcp with HUMANPEN_API_KEY. Apache-2.0.

• vadim-a-yegorov/mcp-ytdlp-transcript: YouTube transcript extraction via yt-dlp with Chrome cookie passthrough — bypasses bot-detection that breaks standard API-based transcript servers.

• TranscriptFetch/mcp-server: Fetch transcripts for YouTube, TikTok and Instagram videos and for podcasts (Spotify, Apple, RSS), transcribing automatically when captions are missing; also search YouTube and list channel or playlist videos. Hosted endpoint at https://transcriptfetch.com/mcp (OAuth). Install: npx transcriptfetch-mcp.

• ArthurBrioche/cadre-video-editor-plugin: Official MIT-licensed local Claude plugin and stdio MCP bridge for Cadre, a proprietary Apple Silicon macOS screen recorder and non-destructive video editor. With Cadre open, it exposes 57 tools for recording, timeline inspection, cuts, cinematic zooms, on-device captions, styling, saving, and MP4 export through an authenticated loopback Agent API. Install: download the v1.0.2 plugin ZIP for Claude Desktop or Cowork, or clone the repository for Claude Code. Requires macOS 13+, Node.js 22, 24, or 26, and Cadre Pro only for export.

• speakai/speakai-mcp: Official hosted server for Speak AI, transcribing and analyzing audio and video recordings in a Speak AI workspace — speech-to-text in 100+ languages with speaker labels, AI-generated summaries and sentiment, call scoring against custom rubrics, clip creation, and captions. Streamable HTTP at https://api.speakai.co/v1/mcp; OAuth 2.1 or Authorization: Bearer <API key>. Also installable via npx @speakai/mcp-server. TypeScript, MIT.

• Synthenova/infiknit-mcp: Local stdio launcher for Infiknit desktop. Agents generate images and videos into a local project and build canvas workflows on the same computer; cloud agents cannot reach it. Install the infiknit.mcpb release bundle or clone the repo and run node plugins/infiknit/bin/infiknit-mcp.cjs while Infiknit is open. Clients: Claude Code, Codex, Cursor, Grok Build, VS Code, Cline, Copilot, and Gemini CLI. Official registry id io.github.Synthenova/infiknit; MIT.

• loudreader/loudkit: Lets agents speak text aloud with a local text-to-speech engine: list the 28 built-in voices (10 languages) and voice profiles cloned with the loudkit CLI, synthesize speech as base64 WAV, FLAC, MP3 or Opus with its duration and a continuation for reading long texts in pieces, and describe the loaded model. Install: uvx --from "loudkit[mcp,hub]" loudkit serve --mcp --checkpoint loudreader/loudr-1 (stdio; no cloud, no API key). Apache-2.0.

• UpRes: AI image and video upscaling to 8K with 6 specialist models (Flare, Prism, Lumen, Mirage, Motion, Motion X) — agents submit upscale jobs, poll status, and download results. Install: npx -y github:auroracapital/upres-cli mcp with UPRES_API_KEY. MIT.

• equalang/equalang-mcp: Translate whole files with layout preservation where supported, translate text and subtitles, and transcribe recordings. Estimate file processing costs before starting paid jobs. An Equalang API key and account credits are required; files are uploaded to the Equalang cloud for processing.

• spicyapi-ai/spicy-mcp: Official SpicyAPI server for image, video and text generation across the SpicyAPI model catalog: browse models, quote prices, create, wait for and retry tasks, upload inputs and fetch outputs. 15 tools. stdio via npx --yes --package=@spicyapi/mcp spicyapi-mcp; SPICY_API_KEY env var. MIT.

• Shorty: Summarize and transcribe videos, audio files, documents and web pages with Shorty. Hosted remote endpoint: https://aishorty.com/api/mcp (Streamable HTTP, OAuth 2.1 with dynamic client registration, no API key). Official MCP Registry: com.aishorty/shorty. Setup for Cursor, VS Code, Cline and Claude Code: https://github.com/DevinoSolutions/mcp-servers/tree/main/shorty

• VoiceLabs: Generate speech from text in your voices and transcribe audio with VoiceLabs. Hosted remote endpoint: https://app.voicelabs.now/api/mcp (Streamable HTTP, OAuth 2.1 with dynamic client registration, no API key). Official MCP Registry: now.voicelabs/voicelabs. Setup for Cursor, VS Code, Cline and Claude Code: https://github.com/DevinoSolutions/mcp-servers/tree/main/voicelabs

• lintlab/pdf-to-markdown: PDF to Markdown converter and PDF text extractor for RAG and AI agents: turns a public PDF URL into page-aware Markdown or plain text with metadata, links and simple tables, and runs OCR on scanned or image-only pages (7 Latin-script languages). Transport: streamable-http, hosted endpoint https://mcp.apify.com?tools=lintlab/pdf-to-markdown. Auth: Apify token required (Authorization: Bearer <token>) or Apify OAuth. Docs: the Actor page; pay per PDF processed and per OCR page, billed to your Apify account. Registry id io.github.lintlab/pdf-to-markdown.

• VideoGen MCP: Official VideoGen server for creating and editing product demos, ecommerce ads, videos, images, voiceovers, music, and avatars, then remixing and exporting projects. Install: npx -y @videogen/mcp with VIDEOGEN_API_KEY, or connect to https://mcp.videogen.io/mcp using Streamable HTTP and OAuth or your VideoGen API key. MIT. Setup and website. Developer docs.

• audiojs/audio: Edit, analyze, and convert local audio files and video soundtracks without ffmpeg. Measure loudness, true peak, and noise floor; check against ACX, podcast, streaming, broadcast, and Netflix specifications; normalize, trim, cut, fade, EQ, denoise, duck, change speed or pitch, mix, split, and detect BPM, key, and chords. Local stdio via npx -y audio --mcp; no authentication. MIT.

• PDF to MD: Convert PDFs and scans to Markdown, export Markdown to PDF, and manage conversion jobs. Five tools upload remote files, create and monitor conversions, read results, and check usage. Hosted Streamable HTTP at https://pdf-to-md.app/mcp; send an account API key as an Authorization bearer token. Docs. Proprietary hosted service.

• CompressGIF: Compress animated GIFs from public HTTPS URLs with compress_gif, monitor jobs with get_compression, and inspect cloud credits with get_usage. Hosted Streamable HTTP endpoint: https://compressgif.net/api/mcp; authenticate with a bearer token or X-API-Key from the account settings. Official MCP Registry: net.compressgif/mcp. Hosted service terms; downloadable adapter source is linked from the documentation.

• Kuro: Official hosted MCP server for creating AI images, video clips, voice-overs, music and editable storyboard films, with reusable recipes and an account-wide media library. Streamable HTTP at https://api.meetkuro.com/mcp; OAuth 2.1 with dynamic client registration and PKCE S256. Install in Claude Code: claude mcp add --transport http kuro https://api.meetkuro.com/mcp. Docs: meetkuro.com/agents.md. Free tier, paid plans and bring-your-own-provider-key generation.

• CreativeClawCo/creative-claw-marketplace: Generate and edit images, video, and audio; reuse characters and brand assets; and create product campaigns, UGC ads, and multi-shot films. Hosted Streamable HTTP endpoint: https://app.creativeclaw.co/mcp with browser-based OAuth. Connection guide. Official MCP Registry: co.creativeclaw/creative-claw. Apache-2.0 plugin package; hosted service uses pay-as-you-go credits.

• liliang-cn/aigui: Lets coding agents draw charts, mermaid diagrams, maths, 3D scenes, molecules and infrastructure topologies whose steps play, as PNGs or a live browser page; fills charts from CSV/JSON files, checks drawings for overlapping or clipped labels, and exports PDF, standalone HTML or GIF. Also ships as a Claude Code and Codex plugin. Install: npm install -g @ai-gui/mcp, then run aigui-mcp. Transport: stdio; no auth. TypeScript/Node; runs locally on macOS, Windows and Linux. MIT.

• VidTL: Official hosted server for VidTL, a browser-based video editor. Edits the user's open editor session live: cut and arrange clips, cut by transcript, generate and style subtitles, clean up audio, add effects, keyframes and motion graphics, pull in stock media, take preview screenshots and export; every change is an ordinary, undoable edit on the timeline. 21 tools. Streamable HTTP at https://vidtl.com/app/mcp; OAuth 2.1 with dynamic client registration and PKCE, no API key. The editor is free; connecting an AI app needs a VidTL Pro plan. Official MCP Registry: com.vidtl/video-editor.

• DenRakEiw/scumble: Operate Scumble, an open-source layered AI inpainting editor for Windows: open images and documents, select by rectangle, mask, point or text description, inpaint the selection into a new layer, work with masks, blend modes, colour match, filter and text layers, take snapshots and undo, look at the result (screenshot), and export PNG, JPEG, WebP, TIFF,...

• Aitho: Official hosted server for Aitho, a presentation rehearsal and delivery app. Turn your own slide deck (PDF or PowerPoint) into a talk: draft and save a speaker script, start a presentation and move between slides, and ask questions answered from your own material. Streamable HTTP at https://present.aitho.app/mcp; OAuth with dynamic client registration and PKCE; sign in to an Aitho account. Free plan, paid plans for more. Official MCP Registry: app.aitho/aitho.

• Shipped-Studio/unsora: Official hosted server for Unsora. Generate images, video, talking-avatar video, music, voiceovers and thumbnails, clone voices, upscale images and video, and cut long videos into short clips, then schedule or publish posts to the user's connected YouTube, TikTok, Instagram, Facebook, LinkedIn, Bluesky, Threads and Pinterest accounts and read post analytics. Results render as interactive previews in clients that support MCP Apps. Streamable HTTP at https://mcp.tryunsora.com/mcp (a clipping-and-posting-only endpoint is at https://mcp.tryunsora.com/scheduler); OAuth with dynamic client registration, or Authorization: Bearer <UNSORA_API_KEY>. Install in Claude Code: claude mcp add --transport http unsora https://mcp.tryunsora.com/mcp. Open source (AGPL-3.0). Docs: https://tryunsora.com/docs/mcp

• makeaivideo-ai/mcp: Official server for MakeAIVideo. Agents can write scripts, estimate credit cost, create finished short-form videos with voiceover, generated or stock scenes, captions, and music, poll jobs, download MP4s, regenerate scenes or voices, and post or schedule videos to connected social accounts. Hosted Streamable HTTP at https://mcp.makeaivideo.ai with OAuth 2.1 or a Bearer API key; local stdio via npx -y @makeaivideo/mcp with MAKEAIVIDEO_API_KEY. 44 tools. MIT. Docs.

• ImgVX AI: Official hosted server for ImgVX AI, free image tools that run in the browser. Agents describe an image job (remove a background, upscale, erase an object, enhance, compress to a file size, resize, crop, convert HEIC, PNG, JPG, WebP or PDF) and get the matching tools with price and a link, in the user's language (40 languages); also lists all 87 tools and pricing. 3 read-only tools; images are never sent to the server. Streamable HTTP at https://imgvx.com/mcp, no authentication. Official MCP Registry: com.imgvx/imgvx-ai. MIT. Docs.

• VAR2.ai: Generate and edit images, create video and music, produce voice and dialog, and work with 3D assets through VAR2's hosted service. Streamable HTTP endpoint: https://www.var2.ai/api/mcp; OAuth sign-in with a VAR2 account, and generation uses account credits. Documented clients include Claude, ChatGPT, Grok, Claude Code, and Cursor. Proprietary hosted service. Setup.

• livepairai/livepair: Hosted LivePair server — community prompt library, live model catalog with a recommender, prompt enhancement, and paid image/video generation plus LLM text; browsing tools are free, generation settles per call in x402 USDC or against prepaid lp_ credit balance. 11 tools. Streamable HTTP at https://livepairai.com/mcp; keyless for browsing, Authorization: Bearer <lp_ key> or x402 for paid tools. MIT.

• Varosity: Official hosted server for Varosity, one account for 60+ generative media models. Compare models and prices, generate video (Veo 3.1, Kling 3.0, Seedance, Runway), images (Nano Banana Pro, GPT Image 2, FLUX), voiceovers, sound effects and music, poll render jobs, and stitch a project's shots into one MP4. 43 tools with read-only and destructive hints. Hosted Streamable HTTP at https://varosity.ai/api/mcp; OAuth 2.1 with dynamic client registration, or a bearer API key; connected apps have a daily spend cap the user sets. Local stdio via npx -y @varosity/mcp-server (MIT). Pay per generation with prepaid credits or bring your own provider keys. Official MCP Registry: io.github.jonkludt/varosity.

• Agent Media: Hosted Streamable HTTP MCP server that makes short drama series (micro-dramas, sitcoms, K-drama style, any language) and other short vertical videos from Claude, ChatGPT or Codex, with the same cast every episode. Endpoint: https://api.agent-media.ai/mcp; OAuth, no API keys.

• gunnargray-dev/youtube-transcripts-mcp: Fetch YouTube transcripts (plain text, timestamped segments or SRT) with video metadata, search YouTube, get chapters, and build extractive summaries with timestamped key points. Install: clone, npm install && npm run build, run node dist/index.js (stdio, no keys); hosted Streamable HTTP version on Apify.

• GeiserX/akou: Local MCP server of akou, an open-source call and meeting recorder that transcribes on the machine. Agents can start and stop a recording, follow the live transcript, search calls and ask questions about them, add notes, name speakers and transcribe audio files. Audio and transcripts are stored on disk and akou has no cloud of its own; questions go to the model provider you choose, which can be a local one. Install: the akou app for macOS, then akou skill install, which registers akou mcp with Claude Code and Codex. Transport: stdio. Auth: none. Docs.

• emirmrkaljevic/mesh2cad: Converts STL/OBJ/PLY meshes into editable STEP CAD by fitting planes, cylinders, spheres and cones, and returns a fidelity report (Hausdorff/RMS deviation, volume error, face types, solid validity). Tools: convert, analyze, report. Install: pip install "mesh2cad[mcp] @ git+https://github.com/emirmrkaljevic/mesh2cad", then run mesh2cad mcp. Transport: stdio; no auth; runs locally (Python, OpenCascade). License: MIT.

• AudioPod AI: Official hosted server for AudioPod's audio AI. Transcribe with speaker labels and SRT/VTT subtitles, split stems (vocals, drums, bass and more), remove noise, separate speakers, convert audio to MIDI or between formats, plus text to speech, voice cloning with consent, voice conversion and music generation; jobs run asynchronously and check_job_status returns download links. 18 tools. Streamable HTTP at https://mcp.audiopod.ai, or https://mcp.audiopod.ai/core for the 9 tools that work on existing audio only; OAuth 2.1 sign-in, or X-API-Key: ap_... for headless use. Free credits on sign-up. Docs: https://docs.audiopod.ai/agents. Official MCP Registry: ai.audiopod/audiopod, ai.audiopod/audiopod-core. Listing repo: MIT.

• elapen/imageloom: Local, offline image studio for agents: generate images from text, edit by instruction or reference picture, upscale (x2 to 8K) and export to PNG, WebP, JPEG or SVG on the user's own machine through stable-diffusion.cpp, with no cloud API. 12 tools: generate_image, edit_image, upscale_image, export_image, get_job, wait_for_job, cancel_job, list_jobs, recent_images, list_models, add_model (from a Hugging Face link) and set_default_model; generation runs as background jobs. Needs the ImageLoom app (or its engine and a model) installed. Transport: stdio; auth: none; no telemetry; Python standard library only; MIT. macOS Apple Silicon is the verified platform, Windows is experimental, and it has so far only been tested with scripted MCP clients (protocol 2025-06-18, 2025-03-26, 2024-11-05), not the real Claude Desktop, Cursor or VS Code. Install: python3 /ABS/PATH/TO/imageloom/core/imageloom_mcp.py (or imageloom mcp; the macOS app bundles it at /Applications/ImageLoom.app/Contents/Resources/core/imageloom_mcp.py). Docs: https://github.com/elapen/imageloom/blob/main/docs/MCP.md

• Shortform Kit: Free remote MCP server for social video: transcribes public Instagram Reels, TikToks and YouTube videos, including videos without captions, reads metadata and captions, downloads media, and returns sourced creator data such as upload specs for 9 platforms, loudness targets and monetization programs. Remote Streamable HTTP endpoint: https://shortformkit.com/mcp; no auth. Docs: https://shortformkit.com/mcp-server/. MIT.

• dare-mcp-server: Generate video and images on Dare (trydare.com) from Claude Code, the Claude desktop app or Cowork, billed to your own Dare credits. Video: Seedance 2.5 and 2.0, Veo 3.1, Kling 3.0 Pro, Hailuo 2.3; images: Nano Banana 2, GPT Image 2, Seedream 5 Pro. 13 tools, including local credit estimates and dry runs. Install: npx dare-mcp-server setup; auth: Dare __client session cookie via setup or DARE_CLIENT_TOKEN. Transport: stdio. Unofficial: uses the same endpoint as the Dare web app. License: MIT.

• AutoShort: Create faceless short videos from a topic, product URL or image; revise scripts, voiceovers, captions and scenes before rendering; poll render status, fetch the MP4 and thumbnail, and inspect credit usage. Hosted Streamable HTTP endpoint: https://autoshort.org/api/mcp with OAuth 2.1 or a Bearer API key. Local stdio alternative: npx -y mcp-server-autoshorts with AUTOSHORT_API_KEY. Generating video spends account credits. MIT for the npm package.

• TubeText TikTok Transcripts, Video Breakdown & Shot Analysis: TikTok transcripts (captions or Whisper AI), quick AI breakdowns and director-level shot analysis. Tools: tiktok-transcript, video-breakdown-ai, shot-by-shot-video-analysis (Apify Actors). Hosted Streamable HTTP at https://mcp.apify.com/?tools=fetch-actor-details,tubetext/tiktok-transcript,tubetext/video-breakdown-ai,tubetext/shot-by-shot-video-analysis; Apify OAuth or Authorization: Bearer <Apify token>. Pay per result on Apify (free monthly credit), failed or empty results not billed. Registry id io.github.LMSlay/tubetext-tiktok-transcripts. MIT.

• TubeText YouTube Transcripts & Channel Analytics: YouTube transcripts from videos, Shorts, playlists or channels, plus channel and Social Blade stats. Tools: youtube-transcript-fast, youtube-channel-analytics, socialblade-scraper, shot-by-shot-video-analysis (Apify Actors). Hosted Streamable HTTP at https://mcp.apify.com/?tools=fetch-actor-details,tubetext/youtube-transcript-fast,tubetext/youtube-channel-analytics,tubetext/socialblade-scraper,tubetext/shot-by-shot-video-analysis; Apify OAuth or Authorization: Bearer <Apify token>. Pay per result on Apify (free monthly credit), failed or empty results not billed. Registry id io.github.LMSlay/tubetext-youtube-transcripts. MIT.

• Cast: Official hosted server for Cast, a mobile-first AI character video app. create_video_link turns one line into a one-tap /make link for a short vertical Seedance video of a character; list_characters, get_prices and get_showcase return the characters, current prices and example videos. Read-only: the server never generates or charges anything, the user opens the link and taps. 4 tools. Hosted Streamable HTTP at https://cast.katechon.technology/mcp; no auth. First 5-second video free, then credit packs. Official MCP Registry: technology.katechon/cast.

• Menivor: Official hosted server for Menivor, an AI short-video generator. Turn a link, product page, HTML or topic into vertical reels with an AI voiceover and word-by-word captions, draft and render product video ads, schedule and list publishes, read performance, and get signed price quotes and wallet balance before spending. 10 tools: create_reel, get_job, render_html, ad_from_url, schedule_publish, list_scheduled, get_performance, generation_quote, wallet_balance, list_media_models. Hosted Streamable HTTP at https://menivor.com/api/mcp; authenticate with a Menivor API key as a bearer token. Proprietary commercial service; first reels free.

• Cutbolt: Local video editing engine for agents: cuts, captions, loudness and export with revisions and undo; needs FFmpeg. Install: release archives or the .mcpb bundle (Windows x64), or the Claude Code plugin (claude plugin install cutbolt --marketplace skulitom/cutbolt-plugin). Transport: stdio. Auth: none (no account, no API key). Official MCP Registry: io.github.skulitom/cutbolt. License: MIT.

• PDFLean: Hosted PDF validation, signature verification, extraction, merging, splitting, compression, OCR, Office conversion, signing and redaction. 17 tools, metered in API credits; job status and credit usage are free. Remote Streamable HTTP at https://mcp.pdflean.com/mcp; OAuth 2.1 with dynamic registration and PKCE. Proprietary hosted service.

• ShortSmash: MCP server + CLI that turns one prompt into a captioned 9:16 short — AI script, neural voiceover in 50 languages, AI-verified stock footage, karaoke captions, one-click publish to 10 platforms. 4 tools: generate_video, get_trends, list_projects, project_status. Install: pip install 'shortsmash[mcp]', then python -m shortsmash.mcp_server. Transport: stdio. Clients: Claude Code, Cursor, Windsurf. MIT.

• AgentCut: Search a directory of AI agent tools for video editors, filtered by app, setup difficulty, price or Resolve edition, and read setup steps, safety notes, license and check date. Hosted Streamable HTTP at https://agentcut.dev/mcp; no auth. MIT. Homepage: https://agentcut.dev.

• Dynamic Document API: Official hosted server for Dynamic Document API. Render PDFs from saved templates, HTML/CSS or Markdown, render PNG, JPEG or WebP images from image templates, merge PDFs, and create e-invoice XML (XRechnung, EN 16931, Factur-X / ZUGFeRD) validated against the official rules. Rendering is deterministic (headless browser, no AI model) and runs in the EU (Germany); free plan with 50 renders a month. 9 tools. Remote endpoint: https://mcp.dynamicdocumentapi.com/mcp. Transport: Streamable HTTP. Auth: OAuth 2.1 (Client ID Metadata Documents or dynamic client registration), or an API key as Bearer token. Local: Docker image franko727/mcp (stdio, API key). Official MCP Registry: com.dynamicdocumentapi/mcp.

• ReelScribe: Official server for ReelScribe. Transcribe Instagram Reels, TikTok, YouTube and Facebook videos by URL, optionally with top comments; submit URL batches or 1–50 reels from an Instagram channel, track jobs, and read, list and search saved transcripts. 9 tools. Hosted Streamable HTTP at https://reelscribe.app/api/mcp (OAuth 2.1 sign-in); local stdio via npx -y reelscribe-mcp-server with REELSCRIBE_API_KEY. 25 free credits, then 1 credit per new video. Docs. Official MCP Registry: io.github.rara-cyber/reelscribe. MIT.

• Lazybird: Official hosted server for Lazybird, an AI voiceover generator. Search 500+ voices in 150+ languages, turn a script into an MP3 voiceover in the voice you choose, design a new voice from a description, and save designed or cloned voices to your library. 5 tools. Hosted Streamable HTTP at https://api.lazybird.app/mcp (OAuth sign-in, nothing to install). Free plan with 10,000 credits a month. Docs. Official MCP Registry: io.github.lazybirdai/lazybird.

• YouTube & TikTok Transcripts MCP Server (heyitsgautham, Apify): Hosted server that gives agents YouTube and TikTok transcripts, YouTube search and YouTube comments. 5 tools: youtube_transcript, youtube_channel_transcripts (a channel or playlist, newest first), youtube_search, youtube_comments and tiktok_transcript. Up to 25 videos per transcript call, 50 search results or 500 comments; optional timestamps, caption language and a character cap per transcript. Reads the captions the site publishes, no speech-to-text. Hosted Streamable HTTP at https://heyitsgautham--video-transcripts-mcp-server.apify.actor/mcp on Apify Standby; auth: Authorization: Bearer <Apify API token>. Paid per result, billed by Apify: $0.003 per YouTube transcript, $0.0015 per TikTok transcript, $0.00024 per search result and $0.00025 per comment, plus the server run's Apify platform usage; videos without captions are not charged. Official MCP Registry: io.github.heyitsgautham/video-transcripts. Source is private.

• Maskbolt: Segment Anything (SAM) on the CPU for agents: cut objects out of images as transparent PNGs, find and count every object, and export masks and SVG/JSON outlines; one Rust binary, no Python, no GPU, nothing uploaded (the model downloads once). 8 tools. Install: release archives or the .mcpb bundle (Windows x64), or the Claude Code plugin (claude plugin marketplace add skulitom/Maskbolt, then claude plugin install maskbolt@maskbolt). Transport: stdio. Auth: none (no account, no API key). Official MCP Registry: io.github.skulitom/maskbolt. License: MIT.

• YouTube Transcript API: Bulk YouTube transcripts with timestamps, flagging human versus auto-generated captions; videos without a transcript are not charged. Hosted Streamable HTTP: https://mcp.apify.com/?tools=catalyst-labs/youtube-transcript-api. Auth: Apify OAuth or bearer API token; pay-per-result hosted service. Connection docs.

• Emero: Hosted MCP server that lets your AI assistant make images, video, music and speech with top models (Kling, Seedance, Nano Banana Pro, GPT Image and more), billed at model cost. A per-brand second brain saves lessons from every generation and reuses them on the next prompt.

• Easy-Peasy.AI: Official hosted server for Easy-Peasy.AI. Run 200+ writing templates, generate and edit images (Nano Banana, FLUX, Seedream, GPT Image), make videos from text or an image (Veo, Kling, Seedance, MiniMax), turn text into speech with 1,000+ voices including cloned voices, and transcribe audio, video files and YouTube, TikTok, Instagram or X links with speaker labels; MCP Apps clients show images and play videos in the chat. 13 tools. Hosted Streamable HTTP at https://easy-peasy.ai/mcp, nothing to install; OAuth 2.1 sign-in or an API key in the x-api-key header. Works with Claude, ChatGPT, Cursor, VS Code, Gemini CLI and other MCP clients. Connecting is free; tools use your plan's credits, and image generation and transcription need a paid plan. Docs. Official MCP Registry: ai.easy-peasy/mcp. The linked repo has client configs and docs; the server code is not public.

• editmamei/editmamei: Photoshop MCP server.

• Giggy.ai MCP: Lets MCP-compatible agents list available voices, generate speech from text, transcribe audio, and check speech-generation status using Giggy.ai.

• RecordHTML: Official hosted server for RecordHTML. Record a public web page, or an HTML document the agent writes (banners, ads, animations, landing pages), to MP4, WebM or GIF in a real browser and get a signed video link. Pages can mark the exact start and end of an animation with trim_start() and trim_end(); each recording reports dropped frames, page errors and failed requests. 5 tools. Hosted Streamable HTTP at https://mcp.recordhtml.com; OAuth 2.1 with dynamic client registration and PKCE, or an API key as a bearer token. Claude Code: claude mcp add --transport http recordhtml https://mcp.recordhtml.com; also works in Claude, ChatGPT, Codex, Cursor and VS Code. Docs: recordhtml.com/docs/mcp. Free plan, then paid plans. Proprietary hosted service. Official MCP Registry: com.recordhtml/mcp.

• varg how to: Official hosted video how-to search for AI agents: 2,000+ video skills, tutorials and FFmpeg recipes. search_how_to takes the person's request in their own words, in any language, and returns ranked cards (agent skills, how-to tutorials, FFmpeg recipes, video formats, prompts and styles) with install commands for Claude Code, Codex, Cursor and Gemini CLI; get_how_to returns the full tutorial as Markdown: the best skill, the 2–4 choices to show the person with example videos, and the steps. 2 read-only tools. Hosted Streamable HTTP at https://api.varg.ai/v2/skills/mcp; no auth, no account or key. Docs: https://varg.ai/how-to/llms.txt. Official MCP Registry: ai.varg/how-to.

• GhostViral: Official hosted server for GhostViral AI, a faceless video maker. The agent writes the script, searches licensed stock footage and inspects thumbnails, sets the scene timeline, makes the voiceover and word-synced captions, then renders a 9:16 Short (15-60 s) or a 16:9 video (5, 8 or 12 min) as MP4, either with FFmpeg on the agent's machine or on GhostViral's server for Shorts up to 60 s. Credits are charged only when the voice is made. Hosted Streamable HTTP at https://ghostviralai.com/mcp; OAuth sign-in, no API key. Works with Claude, ChatGPT, Cursor, VS Code and Claude Code. Free trial credits, then paid plans. Docs. Official MCP Registry: com.ghostviralai/ghostviral. The linked repo has docs and server.json; the server code is not public.

• Toads: Local stdio server for Toads, a Mac app for editing screen recordings. Lets agents tidy up filler words, cut by transcript, add captions, titles and zooms, and export. Runs on your Mac, no API key; needs the Toads website edition (7-day trial, then paid license). Docs: https://toads.io/mcp. MIT connection files; the app is commercial.

• MakeViral: Official hosted server for MakeViral, an AI video and photo maker built around the newest viral trends. Search the latest viral AI video and photo trend templates from TikTok, Instagram Reels and YouTube Shorts and make them from the user's own photos (solo, with a friend or with a pet), or make any AI image or video from a prompt with Veo, Kling, Seedance, Wan, Nano Banana, GPT Image, Seedream or Flux. Check runs, list creations, see the credit balance and get a photo upload link. 14 tools. Hosted Streamable HTTP at https://www.makeviral.ai/mcp; OAuth 2.1 with dynamic client registration (browsing templates needs no sign-in). Docs: https://www.makeviral.ai/developers. Official MCP Registry: ai.makeviral/makeviral.

• Wiro AI: Official MCP server for Wiro AI. Search 3,500+ AI models for images, video, audio, speech, music and text, check each model's parameters and price, run it and get the results back in your assistant. 13 tools. Hosted Streamable HTTP at https://mcp.wiro.ai/v1; OAuth 2.1 browser sign-in, or a Wiro API key as a bearer token. Local stdio: npx -y @wiro-ai/wiro-mcp with WIRO_API_KEY. Connecting is free; each model run is billed from your Wiro wallet. Docs: https://wiro.ai/docs/wiro-mcp-server. Official MCP Registry: ai.wiro/mcp. MIT.

• johnvc/YoutubeTranscripts: Gives agents the timestamped and plain-text transcript of YouTube videos and Shorts in any URL format, with basic video metadata (title, channel, views, duration). Hosted Streamable HTTP at https://mcp.apify.com/?tools=johnvc/YoutubeTranscripts (Apify OAuth, or an Apify API token as Bearer). Pay per video on Apify; failed videos and videos without captions are not charged.

• AppScreen: Official hosted server for AppScreen, a browser editor for App Store and Google Play screenshots. Agents work in the editor tab you have open, so you watch each change: add slides, iPhone, iPad and Android frames, text, logos and your app screenshots (straight from local files), style backgrounds, gradients and patterns, export store-ready PNGs per slide, save and restore works, and make share links. Hosted Streamable HTTP at https://appscreen.pro/mcp; OAuth 2.1 sign-in, no API key. Claude Code: claude mcp add --transport http appscreen https://appscreen.pro/mcp. Works with Claude Code, Cursor, Codex, VS Code and ChatGPT. The editor is free to use; the MCP connection needs AppScreen Pro ($9 one-time). Setup guide.

• GenFlick: Official hosted server for GenFlick, the AI studio for films, series, music videos and ads, built to finish the work, not to make five-second clips. Hosted Streamable HTTP at https://genflick.com/mcp/studios; OAuth sign-in with a GenFlick account (dynamic client registration), no API key. Install: claude mcp add --transport http genflick https://genflick.com/mcp/studios. Proprietary hosted service. Official MCP Registry: io.github.buildngrowsv/genflick.

• inReels: Generate marketing creative: video ads, UGC-style ads with AI actors, static image ads, voiceover and complete faceless videos. 17 tools across 28 image, video and speech models. Read-only tools (list_models, list_voices, quote, get_run_status, ...) are free; generating tools return a run id to poll, and quote gives the exact credit price before anything is spent.