Local Screen Recording Condenser

Shorten Screen Recordings Automatically

Remove long static sections and idle time. Kept sections play at their original speed. Everything runs locally in your browser — nothing is uploaded. The output is a shorter MP4 without audio.

  • No Uploads
  • Local Processing
  • Timeline Preview
  • MP4 Output

Ready.

Error or unexpected result? See troubleshooting

Real Example: 20 Seconds Shortened to 10 Seconds

We made a 20-second test screen recording: 1280 × 720 H.264 MP4 at 30 fps, no audio. It shows a cursor moving for 3 seconds, a 5-second idle pause, a 4-second upload progress bar, a 6-second pause, and a final 2-second cursor move. We condensed it on 31 October 2026 in desktop Chrome 156 with the default settings: Sensitivity 5, static threshold 1.5 s, keep 0.5 s, Balanced quality, and original frame rate.

Original duration
20.0 seconds
Condensed duration
10.0 seconds (10.0 seconds saved, 50% shorter)
Static sections shortened
2
File size
109,115 bytes before, 184,486 bytes after
The actual output: a silent H.264 MP4. A shorter video is not always a smaller file. Here the new encode is larger than the source.

What was kept. The output keeps all three actions: the 3-second opening cursor movement, the full 4-second upload, and the final 2-second cursor move. Each idle pause became 0.5 seconds. Sensitivity 1 and 9 gave the same 10.0-second result on this recording.

What changed. Before 31 October 2026, the tool compared a 9 × 8 pixel fingerprint of each frame. At its old default it cut this recording to about 1 second, and even its strictest setting removed the final cursor move. Always watch the timeline before you export.

Use Condense Video to analyze the recording. Adjust the controls and inspect the timeline, then use Export MP4. The output has no audio. For still images or HDR sources, use the Frame Extractor.

Read the processing and privacy limits. Technical reference: WebCodecs API.

Tested on Six More Recordings

We made six synthetic screen recordings with FFmpeg: 1280 × 720 H.264 MP4 at 30 fps. Each one tests one kind of change. We analyzed them on 31 October 2026 in desktop Chrome 156 on macOS, with static threshold 1.5 s and keep 0.5 s. The table shows the condensed length from the timeline preview. These are six designed cases, not a measure of how often the tool works on real recordings.

Condensed length of six test recordings at two Sensitivity settings
RecordingWhat changesSensitivity 5 (default)Sensitivity 9 (highest)
cursor-move (10 s)A small cursor moves for 6 seconds7.0 s (all movement kept)7.0 s
small-change (10 s)A 40 × 12 pixel status badge changes colour every 2 seconds0.5 s1.0 s
colour-change (10 s)A large card changes fill colour every 2 seconds1.0 s1.0 s
fast-motion (10 s)A large panel moves across the screen the whole time10.0 s (nothing removed)10.0 s
scroll (10 s)A page scrolls for 6 seconds, then stays still6.5 s6.5 s
slides-2min (120 s)A new slide layout every 15 seconds4.0 s (0.5 s per slide)4.0 s

Read the results. The default kept every movement in these recordings: the cursor, the moving panel, and the scroll. Each still slide became 0.5 seconds. Colour-only changes are the weak case: the status badge and card colours changed little in brightness, so most of those changes were shortened. We also tested idle recordings with compression grain and a blinking text cursor. The default shortened them, but Sensitivity 7 and higher kept that idle time. If too much idle time remains, lower the setting. If the tool removes something you need, raise it.

Download the test recordings and try the same settings. Results can differ by browser, device, and decoder.

Why AbeyIO

Kept Sections Play at Normal Speed

Sections that the tool keeps play at their original timing. Very small or colour-only changes can count as static and be shortened. Raise Sensitivity to keep them.

Privacy First

Analysis and encoding run in your browser. Your video is not uploaded.

Preview Before Export

A timeline shows exactly which sections will be shortened and how much time you'll save before you commit.

H.264 MP4 Output

Exports a video-only H.264 MP4. Playback support depends on the receiving player and device.

How It Works

  1. Select a local SDR screen recording, up to 1 GB. It is not uploaded.
  2. Frames are analyzed locally to find static segments.
  3. Preview the condensed timeline and adjust the thresholds.
  4. Click Export MP4 to download a shorter video without audio.

Use Cases

Product Demos

Trim the waiting between clicks so a demo stays tight and watchable.

Bug Reports & QA

Cut idle time from test recordings. Check the timeline to confirm that each step you need is still visible.

Tutorials

Remove loading screens and long pauses. Check that small actions, such as cursor moves, are still in the timeline.

Support Clips

Send a shorter recording. Preview it first to confirm that each step you need is still visible.

What Sets It Apart

Timing-Aware, Not a Slideshow

Unlike frame de-duplicators that stitch stills at a fixed rate, this keeps the original stream timing for active sections and shortens only the stretches that stay still at your Sensitivity setting.

Adjustable Thresholds

Control how similar frames must be to count as static, how long a pause has to last before it's shortened, and how much of it to keep.

Local Processing

Decoding, analysis and encoding run on your device. Run one analysis while online before you rely on offline use. Your video is not uploaded.

FAQ

Does condensing keep animations and transitions smooth?

Kept sections play at their original timing. The tool divides each frame into small blocks and shortens a long section only while no block changes by more than the Sensitivity setting allows. In our tests, the default kept typing, a moving cursor, a thin progress bar, and scrolling, and it shortened idle time with compression grain or a blinking text cursor. Very small colour-only changes, such as a status badge, can still be shortened. Check the timeline before you export.

Is my video uploaded?

No. Analysis and encoding run entirely in your browser with the WebCodecs API. Your video never leaves your device and is never sent to a server.

How does it decide what to remove?

It compares consecutive frames with a perceptual hash. When the picture barely changes for longer than your chosen threshold, that stretch is treated as static and shortened to a short keep duration. Other sections are kept at their original timing.

Which videos work best?

Screen recordings, product demos, tutorials and automated test runs — anything with long static pauses between bursts of activity. Standard (SDR) H.264 or HEVC recordings are supported.

Does it work offline?

Analysis and export run on your device. The processing files load when you first condense a video, so run one analysis while online before you rely on offline use. Opening an unused page without a connection is not guaranteed.

Is the audio kept?

No. The current version exports video without audio. Use it for silent recordings; it is not suitable when narration or sound must be kept.

Is it free?

Yes. It is completely free with no account, no sign-up and no upload.

More Tools

Send feedback