Bitrate beats resolution (usually)
Dropping from 1080p to 720p with the same bitrate often wastes detail. If the frame is already clean, keep resolution and trim bitrate first.
FreeNo WatermarkClient UploadFFmpeg Powered200 MB limit
Crunch video size without killing the quality.
Drop in an MP4 (or almost any common format) and we push it through a tuned FFmpeg recipe. Ideal for squeezing under Discord's 10 MB gate, emailing stubborn clips, or archiving your gameplay library.
Select or Drop Video Files
Uploaded straight to the server. One click away from compressing!
No files yet. Drop or select videos.
Target smaller Discord-friendly clips, trim fat from screen recordings, or just keep your SSD happy. We keep the workflow intentionally boring: choose files, pick a target size, let FFmpeg do its thing.
Why this tool?
- ➤No login, trackers kept minimal, nothing fancy - just compression.
- ➤Reasonable defaults: balances bitrate, scale, and codec tweaks.
- ➤Handles odd resolutions, VFR screen captures, and meme edits.
- ➤200 Megabyte file limit.
- ➤Keeps audio intact unless trimming is needed to hit target.
Quick tip
If you're chasing a specific file size, a lower target + medium quality usually beats cranking quality + guessing. You can always recompress from the source; avoid recompressing a compressed output chain.
How it works
Upload
Your file is streamed to the server. Progress is real, not a fake bar.
Probe
FFmpeg inspects codecs, resolution, bitrate, frame rate, audio layout.
Plan
A target bitrate is derived from your size goal and duration. Safety margins avoid ugly macroblocks.
Crunch
Video is re-encoded (usually H.264) with tuned preset & rate control. Audio may get a gentle bitrate shave if needed.
Deliver
You grab the result. Originals will be deleted shortly after to save disk space.
Fun (nerdy) compression facts
Scroll a little - these pop in as they reach your viewport. Might help you sound smart in voice chat.
Target size math is simple
Desired megabytes × 8192 gives kilobits total. Divide by duration seconds for kbps. We do that, then account for audio overhead.
Compression isn't just shrinking
Encoders exploit temporal redundancy - most frames barely change. Inter frames reference earlier ones instead of storing full pixel grids.
Recompressing hurts
Every lossy pass removes data the next pass can't recover. Keep an original. Each generation adds ringing, banding, and smeared motion.
CRF vs target size
CRF mode targets visual quality; file size floats. Target size mode aims for bytes; quality floats. We blend both by capping extremes.
Containers vs codecs
MP4 is a box. Inside are streams (H.264, AAC, maybe subtitles). Changing the container alone rarely changes size - the codec settings do.
Fast presets cost size
H.264 preset veryfast can inflate bitrate 20-40% vs slow for same quality. We compromise so you aren't waiting forever.