The Official FFmpeg App for Make Is Live

The official FFmpeg Micro app for Make.com is now live in the app directory. If you've been gluing together HTTP modules to call FFmpeg from a Make scenario, you can stop. It's a drag-and-drop module now, the same as Gmail or Google Sheets.

We just shipped a full walkthrough video and a free training hub with five drop-in templates to go with it.

Everything below is free to try on the Make free tier (1,000 ops) and the FFmpeg Micro free tier (100 minutes of processing). No credit card needed on either side.

What's in the app

13 native modules, grouped by what they do:

If you used to call FFmpeg Micro from Make with the HTTP module, you can ditch all of that. The "wait for" modules block until the job finishes, so for any video under five minutes of processing time you can build a scenario as a single straight line: trigger → transcode → save.

For longer jobs, the webhook trigger fires when the job completes, so you split the workflow into a submit scenario and a listen scenario. That pattern is template number two in the training hub.

The five templates

Each one installs into your Make workspace in under a minute. We walk through all of them in the video.

1. Branded videos and thumbnails on autopilot. Drop a video into a Google Drive folder. Get back a 720p watermarked version plus an auto-generated thumbnail in a sibling folder. Built for teams that don't want non-technical people clicking around in video editors. You can swap Google Drive for Notion, Dropbox, Cloudinary, or anything else Make connects to.

2. Async pattern for long videos. Same idea, but split into submit and listen scenarios that talk over a webhook. Use this when your videos take longer than five minutes to process (Make's per-scenario timeout). If you've ever hit a timeout on the HTTP module, this is the pattern you want.

3. Auto-caption every video your team uploads. Drop a video, get back the same video with subtitles burned in. The transcription step uses Whisper; the burn-in step uses the FFmpeg subtitles filter. This scenario is the Make equivalent of our auto-caption guide for n8n.

4. Multi-platform resize (YouTube + Instagram + TikTok). One video in, three videos out. The vertical versions use a blurred horizontal frame as background so you keep the full original content visible inside a 9:16 frame. It looks much better than black bars.

5. Bulk short-form generation from a spreadsheet. Drop quotes into a Google Sheet, drop B-roll clips into a folder. Make picks a random clip, overlays a random quote, generates a video, and marks the row as processed. Run it every 15 minutes and the spreadsheet drains itself. We covered the same idea on the API side in auto-publish 30 YouTube Shorts per week.

The part that's actually new

The five templates are useful, but the section at the end of the video is the bigger deal.

Make has an MCP server. The Make developer hub has an MCP server. FFmpeg Micro has an MCP server. Hook all three into Claude (or Cursor, or any agent platform that speaks MCP) and you can describe a video automation in plain English and have your AI build the scenario for you. No drag-and-drop. No node configuration. No JSON.

In the video we hand Claude a prompt like:

> I want to set up an automation that helps my marketing team brand videos. Whenever a clip lands in this Google Drive folder, process it into a branded version and a thumbnail, and put the results in a sibling folder.

Claude reads the Make docs through the developer-hub MCP, builds the scenario through the Make MCP, runs it, hits an error, reads the execution log, fixes the error, runs it again, and ships a working workflow. Total time: a few minutes. We didn't touch the Make UI.

If you want to see the same MCP setup from the agent side, the deeper write-up is in Build an AI Video Editing Agent with Claude and FFmpeg Micro MCP.

What to do next

If you want to skip the templates and have an AI build your scenario for you, the last 10 minutes of the video walks through the MCP setup step by step.

Why this matters

The old way of using FFmpeg from a no-code platform meant one of three things: you self-hosted FFmpeg behind a custom API, you crammed everything into the HTTP module and hoped you didn't hit a timeout, or you gave up and edited videos by hand.

A native app in Make removes that. Anyone on your team can now build video automations without knowing what FFmpeg is, what a CRF preset is, or how to host a Docker container. They drag, they drop, they're done. And if they don't want to drag and drop either, Claude will build the whole scenario for them.

Both tiers are free to start. There's no reason your team should still be editing videos by hand.