About VidSmaller

VidSmaller is a video compressor that runs in your browser. Drop in an MP4, MOV, MKV or one of 33 supported formats, choose a target, and get a smaller file back — nothing to install, no watermark, and no account needed for your first jobs of the day.

Why it exists

Every place you send a video has a ceiling. Gmail stops at 25 MB. Discord's free tier stops at 10 MB. WhatsApp stops at 16 MB. A 200 MB hero video quietly wrecks your Core Web Vitals. And a phone or screen recording is almost always two to ten times larger than it needs to be, because the recorder optimises for speed, not size.

There were two ways to fix that, and both were annoying:

• Desktop encoders (HandBrake, FFmpeg) do a great job, but a long 4K export pins your CPU for half an hour, and you have to learn what CRF, presets and pixel formats mean before you get a usable file.
• Most web tools stamp a watermark on the output, cap you at 100 MB, or make you sign up before showing you the result.

VidSmaller is the middle path: the controls of a desktop encoder, running on our servers, with the real numbers on screen — and never a watermark.

How it actually works

1. Your browser uploads straight to our storage bucket. The file goes directly from your machine to Cloudflare R2 over a one-time signed URL. It never passes through the web server that renders this page.
2. The encoder pulls it from there, server to server. Our processing provider (FreeConvert) fetches the file from our CDN and runs the FFmpeg job with the settings you picked.
3. You download the result through us. The download is proxied by our server, so the storage layer is never exposed.

The detour through our own bucket is not decoration. Our provider meters the upload step by wall-clock time, so a slow home connection used to cost us — and eventually you — real money. Measured on the same 593 MB file: 6 billed encoder minutes when the browser uploaded directly, 4 when the encoder pulled it from our CDN instead. Roughly a third cheaper, at identical quality.

What we won't do

• No watermark. Not on the free tier, not anywhere.
• We don't train on your footage, resell it, or watch it. No human at VidSmaller opens your video. There is no ML pipeline pointed at it.
• Your files don't linger. The staged upload is deleted the moment a job reaches a final state, with a 24-hour bucket lifecycle rule as a backstop. The compressed result expires on a timer — see the table below.
• No subscription traps. Cancel from your dashboard in one click; it takes effect at the end of the period you already paid for.

The limits, honestly

Two numbers deserve an explanation:

Pro says 1.5 GB, not 5 GB. Our product limit for Pro is 5 GB, but the encoding plan we currently pay for won't accept files that large. Rather than sell you a job we know would be rejected, the app shows the smaller number. It goes up automatically the day we upgrade the encoding plan.

A credit is one billed encoder minute, with a floor of 1. We charge exactly what the encoder charges us. Only the encode is metered — moving your file in and out costs nothing, however long it takes — and the encode is rounded up to a whole minute, so the floor is one credit. That floor covers any 1080p clip up to about seven minutes; the encode only crosses into a second billed minute past that point. A 15-minute meeting recording is 2 credits; a 40-minute 4K file is 5. H.265 costs more because it is roughly 3.8x slower. If a job fails, its credits go straight back to your balance, automatically.

That floor read "3" until 2026-09-08, on the belief that all three pipeline stages were billed. They are not, and we would rather publish the correction than quietly keep the rounding.

This used to read "one credit per minute of source video", priced linearly by duration. Cost is not linear in duration: anything under about seven minutes of 1080p costs the same three machine-minutes. That formula charged 15 credits for a 15-minute meeting recording to buy something we were billed four minutes for. It now tracks the provider's own meter.

What we run on

No mystery stack — if something breaks, this is the list we work through:

Who's behind it

VidSmaller is independent and self-funded. There's no sales team and no investor timeline: the same people who tune the encoder presets answer the support mail. That is also why the pages you're reading list actual limits and actual sub-processors instead of a paragraph about our passion for innovation.

Contact

• Email: support@vidsmaller.com — a human replies, usually within one business day.
• Legal: Privacy Policy · Terms of Service · Refund Policy

Found a file we compress badly? Send it (or a link). Bad cases are the fastest way we improve the presets.