Originally published at ffmpeg-micro.com

YouTube creators spend 10 to 20 minutes per video picking and editing a thumbnail. Multiply that by daily uploads, and you're burning hours on a task that should be automatic. What if you could drop a video URL into Make.com and get a thumbnail generated without opening Photoshop?

This guide walks through a Make.com scenario that takes any video, extracts the best frames using the FFmpeg Micro API, and gives you ready-to-use thumbnails. The whole workflow runs in the background while you focus on creating content.

How the Workflow Works

The scenario follows three steps:

1. Trigger with a video URL (webhook, Google Drive, or manual)
2. Extract frames from key moments using FFmpeg Micro
3. Deliver the thumbnails to Google Drive, Slack, or wherever you need them

Make.com handles the orchestration. FFmpeg Micro handles the video processing. You don't need FFmpeg installed anywhere.

Step 1: Set Up the Make.com Trigger

Create a new scenario in Make.com. For the trigger, you have options:

• Webhook: Send a POST request with the video URL to kick off the workflow programmatically
• Google Drive: Watch a folder for new video uploads
• Manual: Use the "Run once" button during testing

For this tutorial, start with a webhook trigger. Add a "Custom webhook" module and copy the webhook URL. You'll POST to this URL later:

{
  "video_url": "https://storage.googleapis.com/your-bucket/latest-video.mp4"
}

Step 2: Extract Frames with FFmpeg Micro

Add an HTTP module after the trigger. Configure it to call the FFmpeg Micro API to extract a frame from the video.

The trick for thumbnails: extract a frame at a specific timestamp where something visually interesting is happening. For most YouTube videos, the 3 to 5 second mark works well since that's usually past the intro card but into the first visual hook.

HTTP module configuration:

• URL: https://api.ffmpeg-micro.com/v1/transcodes
• Method: POST
• Headers:
  • Authorization: Bearer YOUR_API_KEY
  • Content-Type: application/json

Body:

{
  "inputs": [{ "url": "{{1.video_url}}" }],
  "outputFormat": "mp4",
  "options": [
    { "option": "-ss", "argument": "5" },
    { "option": "-frames:v", "argument": "1" },
    { "option": "-q:v", "argument": "2" }
  ]
}

This tells FFmpeg to seek to the 5-second mark, grab exactly one frame, and save it at high quality. The output is a single-frame video file that you can convert to a JPEG thumbnail.

Step 3: Poll Until the Job Completes

FFmpeg Micro processes jobs asynchronously. Add a second HTTP module to check the job status.

Configuration:

• URL: https://api.ffmpeg-micro.com/v1/transcodes/{{2.id}}
• Method: GET
• Headers: Authorization: Bearer YOUR_API_KEY

Wrap this in a repeater with a 3-second delay. Check if status equals completed. Once it does, move to the download step.

In Make.com, you can use a Router module with a filter: one path continues when status == completed, the other loops back to poll again. Set a max iteration of 20 to avoid infinite loops.

Step 4: Download the Result

Once the job is complete, get the download URL:

HTTP module:

• URL: https://api.ffmpeg-micro.com/v1/transcodes/{{2.id}}/download
• Method: GET
• Headers: Authorization: Bearer YOUR_API_KEY

The response gives you a signed URL valid for 10 minutes. Add another HTTP module to download the actual file from that URL.

Step 5: Deliver the Thumbnail

Now you have the extracted frame as a file. Add a final module to put it where you need it:

• Google Drive: Upload to a "Thumbnails" folder, named with the video title and date
• Slack: Post to a #content channel so your editor can review it
• Airtable: Attach it to a content calendar record

For Google Drive, use the "Upload a File" module. Map the downloaded file from the previous step as the source.

Extracting Multiple Frames

Want options instead of a single frame? Run multiple transcode jobs with different timestamps:

{
  "inputs": [{ "url": "{{1.video_url}}" }],
  "outputFormat": "mp4",
  "options": [
    { "option": "-ss", "argument": "3" },
    { "option": "-frames:v", "argument": "1" },
    { "option": "-q:v", "argument": "2" }
  ]
}

Duplicate the HTTP module three times with timestamps at 3s, 10s, and 30s. This gives you three thumbnail candidates from different moments in the video. Pick the most compelling one.

Common Pitfalls

• Don't skip the polling step. Video processing takes a few seconds. If you try to download immediately after creating the job, you'll get a "not completed" error.
• Use public URLs or presigned URLs for input. The FFmpeg API needs to fetch your video. If it's behind authentication, the job will fail. For private files, use the FFmpeg Micro upload flow to get a gs:// URL first.
• Watch the Make.com operation count. Each HTTP module call is one operation. A scenario with polling can use 5 to 10 operations per run. Factor this into your Make.com plan.
• Set the -q:v flag for JPEG quality. Values range from 2 (best) to 31 (worst). Using 2 gives you a thumbnail sharp enough for YouTube's 1280x720 display.

FAQ

Can I extract frames from YouTube URLs directly?
No. YouTube blocks direct video downloads. You need to upload or host the source video yourself. Google Drive or cloud storage works.

What output format should I use for thumbnails?
The FFmpeg Micro API outputs video formats (mp4, webm, mov). A single-frame MP4 is small enough to use directly, or you can convert it to JPEG in a downstream step using another tool.

How long does frame extraction take?
Usually under 10 seconds. FFmpeg Micro seeks directly to the timestamp without decoding the entire video, so even hour-long videos process quickly.

Can I add text or branding to the thumbnail automatically?
Yes. FFmpeg Micro supports text overlays via the @text-overlay virtual option. You could add a title or logo in the same API call that extracts the frame.

Does this work with other automation platforms besides Make.com?
Absolutely. The FFmpeg Micro API is standard REST, so n8n, Zapier, Pipedream, or any tool that can make HTTP calls works the same way.

Last verified: June 2026