Introducing Video Tools: Compress, Trim & Convert

There is a special kind of frustration that comes from needing a smaller clip right now—and being asked to create an account, upload a two-gigabyte file, and wait for a server on another continent to finish a job your laptop could have done in the background. We built our video suite to end that ritual. Today we are proud to introduce the Ai2Done video tools: compress, trim, format conversion, and more—powered by FFmpeg in the browser, with zero uploads to our infrastructure.

The story behind the launch

Our product team interviewed creators, teachers, and office workers who all described the same workflow: record on a phone, realize the file is enormous, hunt for a “free online converter,” skim a suspicious Terms of Service, and hope for the best. We wanted to offer the opposite experience: drag a file in, see a clear progress bar, download the result, and never wonder where the video went.

Because Ai2Done already runs heavy processing client-side for PDFs and images, extending that philosophy to video was natural—but not trivial. Video codecs are demanding, and browsers impose strict memory and threading constraints. We invested in lazy-loading FFmpeg.wasm, worker-based pipelines so the UI stays responsive, and honest file-size guardrails so users get helpful messages instead of silent failures.

What you can do today

Compress — Shrink file size while keeping quality acceptable for sharing or archiving. You stay in control of the trade-off rather than trusting an opaque cloud preset.

Trim — Cut intros, outros, or mistakes without re-exporting from a desktop editor. Ideal for quick social clips or classroom snippets.

Convert — Move between common containers and codecs when a platform demands a specific format—again, without handing your footage to a server.

Each flow includes clear progress, cancel where safe, and a reset path so experimentation is low-stakes.

Why “no upload” is the headline feature

Video files are uniquely sensitive. They may contain children’s voices, confidential meetings, or unreleased creative work. A privacy-first architecture is not a checkbox; it is the reason users choose us. By running FFmpeg inside WebAssembly, we ensure that the bits you drop into the tab are the bits that stay on your machine—full stop.

Our Go backend still serves HTML, assets, and WASM bundles; it does not need a copy of your media to do its job. That separation is something we intend to preserve as we add more codecs and presets.

Performance and realism

We are upfront: browser video processing is bounded by RAM and CPU. Very large files or exotic codecs may hit limits. We surface those limits in the UI with guidance, not jargon. On the flip side, many everyday tasks—compressing a screen recording, trimming a lecture—complete faster than an upload would have begun.

What is next

This launch is a foundation. We are listening for which workflows matter most: batch queues, GPU hints when available, or tighter integration with our image and PDF suites. If you have been waiting for a video tool that respects your time and your privacy, we hope you will try the new suite and tell us what you think.

Welcome to video on Ai2Done—your files, your device, your pace.