You open ShareX to grab a region, drop an arrow on it, upload it somewhere, and paste a link into a message. Then you switch to a machine where you cannot install it: a locked-down work laptop, a Mac teammate's desktop, a borrowed box with no admin rights. ShareX is Windows-only and needs an installer, so on those machines the whole habit stops. The question is not whether you miss ShareX. It is which of your actual tasks need it at all.
Most ShareX workflows split into two buckets. One is power-user automation: hotkey chains, custom uploader destinations, OCR, GIF recording, workflows that fire on a global shortcut. That bucket is why ShareX exists and a browser tab will not replace it. The other bucket is the everyday job of capturing a screen, marking what matters, and handing someone a link or a file they can open. That second bucket does not need an install anymore.
The capture-annotate-share loop runs in a tab
The core ShareX loop looks like this: capture, annotate with arrows and boxes and text, upload, get a link back. Every step there works in a browser without anything installed. You open a tab, click Capture screen, and the browser's own screen-share prompt grabs the current frame as a still. You crop it, drop numbered pins on specific spots, draw arrows, boxes, and circles, and add short text labels in a few bright colors. On Publish you get a short link anyone can open, plus exports to PDF, Word, and HTML.
That is the part of ShareX most people lean on daily, rebuilt so it runs anywhere a browser does. No admin rights, no installer, no extension. The full comparison lives on the ShareX alternative page, but the short version: if your job is capture and share a marked-up link, the install was never the point.
One difference matters up front. ShareX draws nothing on a live page and neither does a browser capture tool: the markings go onto the captured still, so whoever opens the export sees exactly what you marked. Pins and arrows are baked into the exported image. There is no "the annotation disappeared when I resized the window" problem because the image is frozen at capture.
Where the Windows-only wall actually hurts
The moment you work with people who are not on Windows, ShareX stops being a shared tool. Your designer on a Mac cannot open your ShareX setup. Your client on a phone cannot run it at all. You end up as the single person who owns the capture tool, exporting PNGs and pasting them into email or a chat so everyone else can see them.
A browser flow removes that bottleneck in both directions. On a Mac or Linux box the capture works the same as on Windows. On a phone, where no browser can capture the screen, the flow changes slightly: the person takes a normal screenshot the usual way, then adds it to the review by pasting, dragging, or using Add image. The marking, commenting, and sharing all still work. That means the review link opens on a phone with no app, which is the thing a ShareX PNG in a chat never did cleanly.
Tasks that map cleanly, and the ones that do not
Here is the honest breakdown for the common ShareX jobs:
| ShareX task | Browser tab equivalent |
|---|---|
| Region capture, then arrows and text | Capture screen, crop, pins, arrows, boxes, text labels |
| Upload and copy a share link | Publish gives a short /r/ link anyone can open |
| Hand a bug to a dev with a picture | Owner can send a review to GitHub as an issue |
| GIF or video capture | No. This is still-frame only, not a recorder |
| Custom uploader destinations | No. Output is the link, PDF, Word, HTML, and markdown |
| Hotkey-driven automation and OCR | No. That stays in ShareX |
If you capture video or lean on hotkey workflows, keep ShareX. It is good at that and a browser cannot do it. The split is clean: ShareX for automation and recording, a tab for the capture-and-hand-off work that does not need either.
There is one extra output ShareX does not produce. Every published review is also available as plain markdown, which matters if you hand screens to an AI coding agent. A ShareX PNG is an image an agent has to guess at. A review link an agent opens is served the markdown automatically, pins and comments included. If that is your workflow, the screen capture to markdown path is where to look. For most reviewers it is beside the point, which is why it sits last here, not first.
What it costs and what it does not
You can try the whole capture-annotate-share loop with no password and no credit card. Cobalt asks for an email only when a review is ready, so it can send you the link. The one limit to know: after 7 days a review goes read-only for you, so editing, exports, and Send to GitHub stop. The link you already shared keeps working and nothing is deleted. Pro is $5 a month or $50 a year, keeps every review editable with no cutoff, and cancels any time. Full numbers are on the pricing page.
If you are on a machine where you cannot install ShareX right now, the fastest test is to start a review in a browser tab and run your usual capture-and-mark job through it. If it covers the job, you never needed the installer on that machine. If it does not, you know exactly what ShareX was for.