@agentutility/mcp-mediakit

One endpoint per format. Pay per call.

50 endpoints for PDF, image, video, audio, office, OCR, transcription, watermarking, and format conversion. Whatever the user hands the agent, the agent can handle.

Pricing: pay-per-call in USDC on Base. No subscriptions, no API keys. See per-tool prices below.

Install — Claude Desktop

Edit ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):

{
  "mcpServers": {
    "agentutility-mediakit": {
      "command": "npx",
      "args": ["-y", "@agentutility/mcp-mediakit"],
      "env": { "X402_PRIVATE_KEY": "0xYOUR_PRIVATE_KEY_HEX" }
    }
  }
}

Restart Claude Desktop. 67 tools appear in the tool palette.

Install — Cursor

Add to .cursor/mcp.json:

{
  "mcpServers": {
    "agentutility-mediakit": {
      "command": "npx",
      "args": ["-y", "@agentutility/mcp-mediakit"],
      "env": { "X402_PRIVATE_KEY": "0x..." }
    }
  }
}

Funding

Send any amount of USDC on Base mainnet to the address derived from your X402_PRIVATE_KEY. The MCP server uses it to pay for tool calls automatically.

USDC on Base contract: 0x833589fCD6eDb6E08f4c7C32D4f71b54bdA02913

Tools (67)

| Tool | Description | |---|---| | add-watermark | (0.02 USDC/call) Stamp a text or image watermark onto a PDF, image, or video file. Send a file_url plus either text or image_url, with optional position, opacity, font, rotation, and margin controls, and it runs the CloudConvert watermark task and returns an output_url in the same format as the input. Covers PDFs, common raster formats like PNG/JPG/GIF/WEBP, and video formats like MP4/MOV/WEBM. Use it as a watermarking API, PDF stamping tool, or video/image branding overlay endpoint. | | audio-loudnorm | (0.02 USDC/call) Normalize audio loudness to a target LUFS level for podcasts, YouTube, and Spotify delivery. Send an audio_url plus an optional target_lufs (defaults to -16) and true_peak, and it probes duration, runs two-pass EBU R128 loudness normalization, and returns a hosted MP3 URL with the applied target and duration in seconds. Common presets: -23 for broadcast, -16 for podcasts, -14 for streaming platforms. Use it as a loudness normalizer, LUFS correction API, or audio mastering step before publishing. | | audio-transcribe | (0.01 USDC/call) Transcribes audio to text with whisper-large-v3. Server-side fetches the audio URL (max 25 MB), relays it to Venice's audio/transcriptions endpoint, and returns the transcript with detected language, duration, and per-segment timestamps when response_format='verbose_json' (default). Also supports raw text, SRT, and VTT outputs. Use it as a speech-to-text or multi-language ASR endpoint with OpenAI Whisper API compatibility. | | compress-pdf | (0.005 USDC/call) Shrink a PDF's file size without manually tuning compression settings. Send a pdf_url and an optional profile: web (default), print, archive, mrc for scanned documents, or max. It runs CloudConvert's optimize task and returns a pdf_url along with original_size_bytes, compressed_size_bytes, and saved_percent. Use it as a PDF compressor, PDF size reducer, or PDF optimization API before emailing or uploading large documents. | | convert-html-to-markdown | (0.005 USDC/call) Convert HTML to Markdown. Strips nav, scripts, ads, and other boilerplate. Preserves headings, lists, tables, code blocks, links, and images. Accepts raw HTML or a URL. Returns clean Markdown ideal for LLM context windows or RAG ingestion. | | convert-pdf | (0.0025 USDC/call) Converts PDFs to Markdown, HTML, JSON, or structured text with the Datalab Marker AI pipeline, preserving headings, tables, equations, and multi-column flow. Note: this endpoint outputs text-format conversions only. For PDF to JPG/PNG use pdf-to-jpg, for PDF merge use pdf-merge, for PDF split use pdf-split, for PDF compression use compress-pdf, and for PDF to Word route through this then a markdown to docx step. 30 pages max. Use it as a general PDF converter. | | csv-to-ics | (0.01 USDC/call) Converts a CSV of events into an RFC 5545 compliant ICS calendar file (VCALENDAR/VEVENT) for Google Calendar, Outlook, and Apple Calendar import. Auto-detects column mapping (summary, date, time, location, description) or accepts an explicit mapping; handles all-day and timed events, configurable default duration, and custom delimiters. Up to 1000 rows. Use it as a CSV to iCal converter, calendar file generator, or .ics builder. | | csv-to-jsonl | (0.02 USDC/call) Converts CSV or TSV data into JSON, JSONL/NDJSON, or column-oriented arrays. RFC 4180 parser with type inference (booleans, integers, floats, ISO dates, null tokens) and configurable delimiter, quote char, header presence, column rename, drop columns, and whitespace trim. Outputs newline-delimited JSON (jsonl), a JSON array (json), or column-oriented arrays (ndarray). Use it as a CSV to JSON, CSV to NDJSON, TSV to JSON, or spreadsheet to JSON converter, data pipeline preprocessor, or LLM training-data prep step. | | doc-to-json | (0.10 USDC/call) Converts any document (PDF, DOCX, PPT, XLSX, or image) into structured JSON matching a caller-supplied schema. The file is converted to plain text via CloudConvert, then Morpheus mistral-31-24b runs with strict json_schema response_format to force the output shape. OCR-aware for images and scanned PDFs. Use it as a file parser with schema, invoice extractor, resume parser, contract extractor, RFP parser, receipt extractor, or form data extraction tool. | | excel-to-csv | (0.005 USDC/call) Converts Excel workbooks (.xlsx / .xls) to CSV, TSV, or JSON with multi-sheet handling, returning each sheet by name. Adjacent to the 'convert excel to google sheets' demand cluster. Use it as an Excel to CSV or spreadsheet converter. | | excel-to-google-sheets | (0.005 USDC/call) Converts Excel files for import into Google Sheets. Outputs CSV that imports directly via File > Import > Upload (or paste-into-cells), with multi-sheet handling, encoding control, and quote style. Same handler as excel-to-csv / xlsx-to-csv under a Google-Sheets-named slug. Use it as an XLSX to Google Sheets, Numbers to Google Sheets, or Excel to gsheet spreadsheet import tool. | | extract-tables | (0.10 USDC/call) Detects and extracts every table from a PDF document, returning structured JSON or CSV per table. 30 pages max via Datalab Marker. Use it as a table extractor, PDF-to-CSV converter, or spreadsheet-from-PDF tool. | | html-markdown | (0.005 USDC/call) Converts raw HTML or a URL into clean Markdown, preserving headings, lists, tables, code blocks, links, and images. Use it as an HTML to Markdown API, webpage to Markdown converter, or clean page text extractor. | | html-to-markdown | (0.005 USDC/call) Convert HTML to Markdown. Strips nav, scripts, ads, and other boilerplate. Preserves headings, lists, tables, code blocks, links, and images. Accepts raw HTML or a URL — when given a URL the server fetches the page and runs the converter on the body. Returns clean Markdown plus title and char counts. Ideal for LLM context windows, RAG ingestion, and content archival. | | html-to-pdf | (0.08 USDC/call) Render a live webpage as a PDF document. Send a url plus optional page_format, orientation, margins, and a wait_until condition for JS-heavy pages, and it captures the rendered page through CloudConvert and returns a pdf_url with the file size in bytes. Use it as a URL to PDF converter, webpage-to-PDF capture tool, or website screenshot to PDF generator for archiving or reporting pages. | | image-convert | (0.01 USDC/call) Convert an image from one format to another. Send an image_url and output_format (PNG, JPG, WEBP, AVIF, GIF, BMP, TIFF, ICO, HEIC, HEIF, or PSD), with optional width, height, fit, and quality for resizing, and it returns an image_url in the new format with the resulting file_size_bytes. Handles SVG and RAW as inputs too. Use it as an image format converter, HEIC to JPG tool, or image resize and compress API. | | image-format-convert | (0.01 USDC/call) Image converter. Convert any image between PNG, JPG, WEBP, AVIF, GIF, BMP, TIFF, ICO, HEIC, HEIF, PSD, and SVG. Optional resize and quality. CloudConvert engine. Same backend as image-convert under a more search-friendly slug. | | image-translate | (0.02 USDC/call) Image translator. Translate text in any image — signs, menus, screenshots, photos of documents — by running vision-OCR plus Venice translation. Returns the detected source text plus the target-language translation. Optional preserve_layout mode keeps the original spatial arrangement. | | image-upscale | (0.02 USDC/call) Upscales an image 2x or 4x via Venice's image/upscale endpoint (default model: venice-sd35). Returns a permanent fal-hosted URL. Use it as a super-resolution tool to sharpen or enlarge an image without loss. | | image-watermark | (0.02 USDC/call) Image watermark / add text or logo watermark to image files. Also supports PDFs and videos through the same CloudConvert-backed watermark handler. Configurable opacity, position, rotation, font, and margin. | | json-yaml | (0.005 USDC/call) Converts between JSON and YAML in either direction. Auto-detects input format from the first non-whitespace byte ('{'/'[' = JSON, otherwise YAML), parses, and re-serializes in the requested target format; pretty-print and indent are configurable. Use it as a format transcoder, config-file translator, Helm values prepper, Kubernetes manifest converter, GitHub Actions workflow YAML helper, or round-trip parser. | | logo-detect | (0.03 USDC/call) Brand logo detection / brand recognition in images. Venice qwen3-5-9b vision LLM. Returns detected brands with confidence, location, evidence (wordmark/logomark/lockup/color_scheme), element_type, and supports hint_brands. | | merge-pdf | (0.01 USDC/call) Combines two to fifty separate PDF files into one, taking each source as a URL and stitching them together in the order supplied, bookmarks intact. Send pdf_urls (an array of links) and an optional filename for the output; get back pdf_url pointing to the merged file and merged_count with the number of inputs combined. Same backend as pdf-merge. Use it as a PDF merger to concatenate PDF files, join multiple PDFs into one document, or assemble multi-file submissions into a single deliverable. | | movie-database | (0.01 USDC/call) Finds movies or TV shows by title, with optional year and region, and returns release year, poster, overview, and language. Uses TMDB when TMDB_API_KEY is configured (adding movie ratings, popularity, and an optional streaming provider summary) and falls back to no-key Wikidata metadata otherwise. Use it as a movie database API, TV database API, film lookup, title metadata search, or TMDB / Wikidata movie lookup. | | movie-database-api | (0.01 USDC/call) Searches movies and TV shows by title and optional year, returning release date, rating, popularity, overview, poster URLs, TMDB links, and optional provider summary. Use it as a movie database API, TV database API, or TMDB metadata lookup. | | movie-info | (0.01 USDC/call) Looks up movie and TV metadata: title, release year, rating, overview, poster, and optional streaming providers. Same TMDB-backed response as movie-database with a compact search slug for agents. Use it as a movie info API, film metadata lookup, or TV metadata lookup. | | mp4-to-mp3 | (0.02 USDC/call) Converts MP4, MOV, WebM, MKV, AVI, M4V, and FLV video files to MP3 via CloudConvert, with selectable bitrate (96/128/192 kbps). 60-min / 500MB max. Use it as an MP4 to MP3 converter, video to audio converter, or audio extractor API. | | mp4-to-mp3-api | (0.02 USDC/call) Converts MP4, MOV, WEBM, MKV, AVI, M4V, or FLV URLs into hosted MP3 output with selectable 96, 128, or 192 kbps bitrate. CloudConvert backend, 60-min / 500MB max. Use it as an MP4 to MP3 API, video to MP3 converter, or extract-audio-from-video endpoint. | | mp4-to-mp3-converter | (0.02 USDC/call) Converts MP4 video URLs into hosted MP3 files with selectable bitrate; also accepts MOV, WEBM, MKV, AVI, M4V, or FLV URLs. Same CloudConvert-backed handler as mp4-to-mp3. Use it as an MP4 to MP3 converter API, to extract audio from MP4, or for video-to-MP3 conversion. | | ocr | (0.0025 USDC/call) Runs OCR on scanned PDFs and image-based documents, returning clean Markdown or plain text. The Datalab Marker engine preserves layout, tables, and math. 30 pages max. Use it as an OCR API, optical character recognition service, scanned document extractor, or image-PDF to text and scanned PDF OCR tool. | | office-to-pdf | (0.05 USDC/call) Converts office documents and other file formats into PDF from a URL, no upload step required. Accepts DOCX, DOC, XLSX, XLS, PPTX, PPT, ODT, ODS, ODP, RTF, TXT, CSV, TSV, EPUB, Markdown, HTML, and Apple Pages, Numbers, and Keynote files, auto-detecting the source format from the file extension. Send file_url and get back pdf_url plus source_format, run through the CloudConvert engine. Use it as an office to PDF converter, document conversion API, DOCX-to-PDF tool, or spreadsheet and slide-deck to PDF service. | | pdf-compress | (0.005 USDC/call) Shrinks PDFs to a smaller size for email or the web, with three quality levels: ebook (lowest, web-quality), printer (medium), prepress (highest, archival). CloudConvert engine. Use it as a PDF compressor or PDF size reducer. | | pdf-extract-tables | (0.10 USDC/call) Extracts every table from a PDF, digital or scanned, and returns row-by-column text matrices page-by-page. AI + OCR pipeline with optional cell bounding boxes for downstream layout reconstruction and an optional page_range filter ('1-5', '3', '1,3,5'). Handles merged headers, multi-page financial statements, balance sheets, lab results, scanned reports. 30 pages max. Sibling of pdf-to-markdown using the same Datalab backend, but pre-parsed to tables only. Use it as a PDF table extractor, scanned-table parser, financial-table OCR, multi-page table consolidator, or Datalab Marker tables endpoint. | | pdf-merge | (0.01 USDC/call) Joins multiple PDF files, hosted at URLs, into a single ordered document, so you don't have to merge them by hand in a desktop PDF editor. Send pdf_urls as an array of 2 to 50 links in the order you want them combined, plus an optional filename; the merge engine combines them and preserves each source's bookmarks. The response returns pdf_url for the combined file and merged_count confirming how many sources it used. Use it as a PDF merger, PDF combiner, PDF concatenator, or document-joining API for stitching reports, contracts, or scanned pages into one file. | | pdf-parser-api | (0.0025 USDC/call) Parses a public PDF URL into Markdown, HTML, or JSON blocks with layout-aware text, headings, tables, and equations. Datalab Marker backend, 30 pages max. Use it as a PDF parser API, PDF content extractor, or scanned PDF OCR API. | | pdf-split | (0.04 USDC/call) Split a PDF into multiple files by page range or one file per page. Send a pdf_url and either ranges (like ['1-3','5','7-end']) for custom chunks, or split_each_page:true to explode every page into its own PDF. Runs on the CloudConvert split task and returns a pdfs array with a hosted url and range label per output file, plus an output_count. Use it as a PDF splitter, PDF page extractor, PDF page range downloader, or PDF-to-multiple-files converter for report chunking, contract page pulls, or batch page exports. | | pdf-table-extract | (0.10 USDC/call) Extracts tables from digital or scanned PDFs, returning row/column matrices, CSV output, page numbers, and optional cell boxes. Same Datalab Marker-backed handler as pdf-extract-tables. Use it as a PDF table extract API, to extract tables from PDF, or as a PDF to CSV table parser. | | pdf-table-extractor | (0.10 USDC/call) Finds tables in digital or scanned PDFs and returns row-by-column matrices, page numbers, and optional cell bounding boxes. Datalab Marker backend, 30 pages max. Use it as a PDF table extractor or scanned-table OCR. | | pdf-text-extractor | (0.0025 USDC/call) Extracts clean Markdown, HTML, or structured JSON from digital or scanned PDFs while preserving reading order, tables, and equations. Datalab Marker backend, 30 pages max. Use it as a PDF text extractor, PDF to text API, or OCR PDF reader. | | pdf-to-jpg | (0.10 USDC/call) Convert a PDF into page images for previews, thumbnails, or archiving. Send a pdf_url with an optional format (jpg, png, or webp), quality, and pixel_density (36-600 DPI), and it rasterizes every page and returns one image URL per page along with the total page_count. Use it as a PDF to JPG converter, PDF page rasterizer, or PDF thumbnail generator when you need images instead of a PDF viewer. | | pdf-to-markdown | (0.0025 USDC/call) Converts digital or scanned PDFs to clean Markdown with AI-powered, layout-aware extraction on the Datalab Marker engine. Preserves headings, tables, equations (LaTeX), bulleted lists, and multi-column flow; outputs Markdown (default), HTML, or structured JSON with per-page blocks. 30 pages max. Use it as a PDF parser, PDF to text converter, OCR PDF reader, extract-tables-from-PDF tool, equation-aware PDF parser, scanned-PDF OCR, or PDF data extractor. | | pdf-to-markdown-api | (0.0025 USDC/call) Converts a public PDF URL into clean Markdown, HTML, or structured JSON while preserving headings, tables, equations, and reading order. Datalab Marker backend, 30 pages max. Use it as a PDF to Markdown API, PDF parser API, PDF to text API, or scanned PDF OCR API. | | pdf-to-text | (0.0025 USDC/call) Extracts text from digital or scanned PDFs, preserving reading order across multi-column layouts with an AI + OCR pipeline (Datalab Marker). Recognizes scanned pages and returns Markdown by default (clean text with structure) or HTML / JSON. 30 pages max. Works as a PDF to plain text converter, pdftotext or pdf2txt replacement, PDF text extractor, scanned PDF OCR, or read/parse PDF content extractor. | | pdf-to-text-api | (0.0025 USDC/call) Extracts text from digital and scanned PDFs as Markdown, plain text, HTML, or JSON with layout-aware reading order. Same Datalab Marker-backed handler as pdf-to-text. Use it as a PDF to text API, PDF text extractor, or OCR PDF reader. | | pdf-watermark | (0.02 USDC/call) Overlays a text or image watermark on PDFs, PNG/JPG/GIF images, or MP4/MOV/WEBM videos. Configurable position, opacity, font, rotation, and margin. CloudConvert engine. Use it as a PDF watermark, image watermark, or video watermark tool. | | pdf2md | (0.20 USDC/call) Convert a hosted PDF into clean Markdown, HTML, or structured JSON by sending its pdf_url. Runs on Datalab's Marker OCR engine, which is layout-aware, so multi-column pages, tables, and equations come out readable instead of scrambled the way naive text extraction leaves them. The response includes the converted output plus page_count and the source URL; PDFs over 30 pages are rejected up front so you know immediately instead of waiting on a doomed job. Use it as a PDF to Markdown converter, PDF OCR API, or document text extraction tool for contracts, papers, and scanned reports. | | receipt-ocr | (0.01 USDC/call) Receipt OCR. Reads any receipt photo and returns a structured JSON object with vendor, address, date, line items (qty / unit_price / total), subtotal, tax, tip, total, and payment method. Vision-LLM powered. Same backend as receipt-parser under a clearer slug for expense + accounting + reimbursement workflows. | | receipt-parser | (0.01 USDC/call) Receipt parser. Extracts structured JSON from any receipt image — vendor, address, date, line items (qty / unit_price / total), subtotal, tax, tip, grand total, and payment method. Vision LLM powered. Useful for expense automation, accounting integrations, and travel-claim flows. | | speaker-diarize | (0.10 USDC/call) Transcribe audio or video and label who said what. Send a media_url (optionally language and num_speakers) and it runs Whisper v3 speech-to-text with speaker diarization, returning an utterances array grouped by speaker with start/end timestamps and text, plus per-speaker stats (utterance count, seconds spoken, word count) and total duration_seconds. Files up to 60 minutes are supported. Use it as a speaker diarization API, who-said-what transcription tool, multi-speaker transcript generator, or meeting/interview transcription service for calls, podcasts, and recorded interviews. | | split-pdf | (0.04 USDC/call) Split a PDF into multiple files by page range or one PDF per page. Send a pdf_url and either ranges (e.g. ['1-3','5','7-end']) to cut custom chunks, or split_each_page:true to break every page out on its own. Backed by the CloudConvert split task, it returns a pdfs array of hosted file URLs tagged with the range they cover, plus an output_count. Use it as a PDF splitter, PDF page extractor, split-PDF-by-range tool, or PDF-to-multiple-files converter for pulling exhibits, invoices, or chapters out of a longer document. | | subtitles | (0.08 USDC/call) Generate ready-to-use subtitle files from a video or audio URL. Send media_url with an optional language, output format of srt or vtt, task of transcribe or translate, and a max_chars_per_line setting, and get back a properly timestamped, word-wrapped subtitles string you can drop straight into VLC, Premiere, or FFmpeg. Runs on Whisper v3 for the underlying transcription and checks media duration first, rejecting anything over 60 minutes. Use it as a subtitle generator, closed-caption creator, or SRT/VTT converter for video editing, accessibility captions, or localization workflows. | | transcribe | (0.10 USDC/call) Convert spoken audio or video into text. Send a media_url (mp3, mp4, mpeg, mpga, m4a, wav, or webm) with optional language and task settings, and get back a text transcript. Runs on Whisper v3, auto-detects 90+ languages, supports a translate-to-English mode instead of same-language transcription, and checks media duration up front, rejecting anything over 60 minutes or 500MB before spending time on the job. Use it as a video transcription API, speech-to-text converter, or audio transcript generator for podcasts, interviews, meetings, or any spoken-word source you need in text form. For speaker labels, use the speaker-diarize endpoint. | | upscale-image | (0.02 USDC/call) Upscales images 2x or 4x with AI super-resolution, best for photos and illustrations. Venice image/upscale backend (venice-sd35), re-hosted for a permanent URL. Use it as an AI image upscaler or image enlarger. | | url-to-markdown | (0.005 USDC/call) Fetches a public URL, strips boilerplate, and returns clean Markdown plus title and character counts. Use it as a URL to Markdown converter, webpage to Markdown tool, or article cleanup API for pulling readable text out of blog posts, docs pages, and news articles before feeding them to an LLM. Send a url; get back source_url, title, markdown, markdown_chars, and html_chars so you can gauge how much boilerplate got trimmed. | | video-summarize | (0.10 USDC/call) Summarizes videos, podcasts, and lectures in one call: Whisper v3 transcribes, then Mistral summarizes. 5 styles (tldr, bullets, paragraph, executive, chapters); returns summary + transcript; 60 min max. Use it as a video summarizer, podcast summarizer, or lecture notes generator. | | video-thumbnail | (0.03 USDC/call) Pull a single still frame out of a video without transcoding the whole file. Send a video_url and pick frame_type: first, middle, or last, and get back a hosted JPG with its width and height. Runs on fal.ai's ffmpeg extract-frame pipeline, so it's fast because it only decodes the frame it needs. Use it as a video thumbnail generator, video frame extractor, video-to-image converter, or preview-image API for video libraries, CMS thumbnails, and content moderation previews. | | video-to-audio | (0.02 USDC/call) Video to audio extractor / video to audio converter. Extract MP3 audio track from any video URL (MP4, MOV, WEBM, MKV, AVI, M4V, FLV). Selectable bitrate (96/128/192 kbps). Useful for podcast extraction, audio archival, transcription pre-processing. 60-min / 500MB max. CloudConvert backend. | | video-to-mp3 | (0.02 USDC/call) Video to MP3 / extract MP3 audio from video URL. Converts MP4, MOV, WEBM, MKV, AVI, M4V, or FLV into an MP3 file for transcription, podcast clips, and media archival. CloudConvert backend, 60-min / 500MB max. | | video-to-subtitles | (0.02 USDC/call) Generates subtitles from video with Whisper v3, word-wrapped and ready for VLC / Premiere / FFmpeg. Auto-detects language and can translate to English. Use it as a video subtitle generator, auto-subtitle and closed captions tool, SRT generator, VTT generator, video CC endpoint, or accessibility captions source. | | video-to-text | (0.10 USDC/call) Transcribe any video URL to text with Whisper v3 large. Audio is extracted internally. Auto-detects 90+ languages, offers a translate-to-English mode, and handles files up to 60 minutes / 500MB. Use it as a video transcription API, video speech-to-text tool, or video ASR endpoint. For speaker labels, use the speaker-diarize endpoint. | | video-transcribe | (0.10 USDC/call) Transcribes any video URL to text with Whisper v3 large. Auto-detects 90+ languages and offers a translate-to-English mode. 60-min / 500MB max. Same backend as video-to-text under a clearer slug. Use it for video transcription, video to audio transcription, or video-to-text. For speaker labels, use the speaker-diarize endpoint. | | video-trim | (0.02 USDC/call) Cuts a clip out of a source video by start and end point, so you don't need a video editor to pull a highlight or shorten a file. Send video_url plus trim_start and either trim_end or duration, each as HH:MM:SS, MM:SS, or plain seconds; an optional output_format lets you change the container on the way out. The trimmed segment is re-encoded with an x264 pipeline, preserving audio, and the response returns the new video_url with the resolved trim_start and trim_end. Use it as a video trimmer, video cutter, clip extraction API, or video editing tool for shortening uploads and pulling highlight clips. | | watermark | (0.02 USDC/call) Adds a text or image watermark overlay to PDFs, PNG/JPG/GIF images, and MP4/MOV/WEBM videos. CloudConvert engine. Configurable position, opacity, font, rotation, and margin. Use it for PDF, image, or video watermarking. | | watermark-pdf | (0.02 USDC/call) Add watermark to PDF. Text or image overlay on PDFs, PNG / JPG / GIF, or MP4 / MOV / WEBM with configurable position, opacity, font, rotation, and margin. Same backend as watermark / pdf-watermark under a clearer search slug. CloudConvert engine. | | xlsx-to-csv | (0.005 USDC/call) Converts any sheet of an .xlsx, .xlsm, .xls, or .ods workbook to CSV, with sheet selection, encoding, and quote style controls. Powered by CloudConvert. Use it as an Excel to CSV, XLSX to CSV, Numbers to CSV, or spreadsheet to CSV converter. | | xml-to-word | (0.05 USDC/call) Converts an XML document into a Microsoft Word (.docx) file for delivery to systems that need Word format, not raw XML. Send xml_url, a public HTTPS link, or inline xml text plus an optional title; the endpoint renders the XML into structured HTML that preserves element nesting as headings and lists, converts that HTML to DOCX, and returns a downloadable docx_url along with file size and character counts. Useful for legal and finance pipelines that emit XML reports but need Word deliverables, regulator-format compliance workflows, and agent-led document generation. Use it as an XML to Word converter, XML to DOCX API, or XML document formatter for downstream Word delivery. | | youtube-transcript | (0.01 USDC/call) Fetch the transcript of any YouTube video. Pulls auto-generated or manual captions and returns full text plus per-segment {start, duration, text}, with an optional language pick. Backed by Supadata's transcript pipeline server-side — no caller key required; returns a clear 404 when a video has no transcript. Use it as a YouTube closed-caption fetcher or video subtitles puller. |

How it works

1. Agent calls a tool (e.g. add-watermark).
2. MCP server POSTs to https://x402.agentutility.ai/add-watermark.
3. The endpoint responds HTTP 402 with payment instructions.
4. The MCP server signs an EIP-3009 USDC transfer authorization with X402_PRIVATE_KEY and retries.
5. CDP facilitator settles on Base.
6. The endpoint returns the actual response.

The agent never sees the payment flow — it just gets the result.

Links

• Cluster overview: https://agentutility.ai/mediakit/
• All MCP packages: https://mcp.agentutility.ai/
• Source: https://github.com/rooz21/x402/tree/main/packages/mcp-mediakit

Version: 0.11.11 · License: MIT