Frequently Asked Questions

SnapSight captures periodic screenshots of your screen using your browser's Screen Capture API — no installation required. A local AI model (Florence-2) runs entirely on your device, classifies what you're working on, and assigns that time to the right project. No timers to start or stop, no manual logging. You just work, and SnapSight builds your timesheet in the background. Learn how it works in detail →

Yes. The free plan includes 10 screenshots per day, AI project classification, dashboard charts, CSV export, and PWA offline support — no credit card, no trial countdown. The Pro plan at $5/month adds unlimited screenshots, auto-capture, invoice-ready Project Time Reports, and weekly push notifications. The free plan is genuinely usable, not a bait-and-switch.

SnapSight runs in any modern desktop browser that supports the Screen Capture API — Chrome, Edge, and Firefox on Windows and macOS. Mobile devices are not supported because mobile browsers do not allow screen capture. No app installation is required; it works as a Progressive Web App (PWA) directly from your browser.

Open the SnapSight app, go to Projects, and create a project with a name and keyword rules. Keyword rules tell the AI which window titles or app names belong to that project — for example, "Figma", "VSCode", or your client's name. Once set up, the AI automatically routes captured screenshots to the right project. The setup takes about two minutes. See our guide on tracking billable hours across multiple projects.

Yes. SnapSight is a PWA with a service worker that caches the app shell for offline use. The Florence-2 AI model runs locally in your browser using WebAssembly, so classification works without an internet connection. Your time data is stored in your browser's IndexedDB. Dashboard syncing happens automatically when you're back online.

No. Every screenshot is processed locally by the Florence-2 model running in your browser. The raw image is classified and discarded — it never leaves your device, and it's never sent to any server. SnapSight only stores the classification result (project label, timestamp, duration) in your browser's local storage. There are zero cloud AI calls. This is the core reason remote workers choose SnapSight over surveillance tools.

No. SnapSight is a self-tracking tool, not an employer surveillance product. Screenshots are captured and processed locally on your machine. No one else — not your employer, not SnapSight — can see your screen activity. The data lives in your browser and stays there. Read more about time tracking without surveillance.

Classification results (project label, timestamp, duration in seconds) are stored in your browser's IndexedDB. Aggregated totals sync to your SnapSight account to power the dashboard charts. Raw screenshots are never stored anywhere — they're classified and immediately discarded. Your account data is protected with industry-standard encryption at rest.

The AI classification layer uses Florence-2, an open-source vision model from Microsoft, running via Transformers.js. The SnapSight application itself is not open source, but the local AI component is fully auditable — you can verify it makes no network calls using your browser's developer tools network panel.

SnapSight runs Florence-2, a compact vision model, directly in your browser using WebAssembly. When a screenshot is captured, the model analyzes the image alongside the active window title to determine which project it belongs to. Your project keyword rules act as a routing layer on top of the AI output — if the model sees "VSCode" and you've mapped it to a client project, the time logs there automatically. Classification typically takes under a second. See how this compares to other automatic time tracking approaches.

Yes — multi-project tracking is the core use case. Create a project for each client and define keyword rules that map your apps and window titles to the right project. SnapSight separates and totals time per project automatically. The dashboard shows a breakdown by project, and the Project Time Report (Pro) exports per-project hours ready for invoicing. Freelancers and agencies use this daily.

Free plan users can export raw classification data as CSV or JSON from the dashboard. Pro users get the Project Time Report — a clean, invoice-ready CSV with per-project hour totals, formatted for direct attachment to client invoices or import into billing tools. You can also use our timesheet templates or the billable hours calculator alongside your exports. See our guide on turning time tracking data into client invoices.

Manual timers require you to remember to start and stop them every time you switch tasks. SnapSight runs automatically in the background — no timers, no mental overhead, no forgotten sessions. Research shows manual time tracking underreports billable hours by 20–30% because people forget to log or round down. SnapSight captures every working minute without any action on your part. See our full comparison with Toggl →

Not natively yet — SnapSight exports invoice-ready CSVs that you import into your invoicing tool of choice (FreshBooks, QuickBooks, Wave, HoneyBook, etc.). Native integrations are on the roadmap. In the meantime, the Project Time Report CSV is formatted to minimize import friction, and our timesheet templates cover common invoice formats.

The free plan includes 10 screenshots per day, local AI classification with Florence-2, daily/weekly/monthly dashboard charts, project management, CSV and JSON export, and PWA offline support. No credit card required and no time limit. It's a fully functional tier, not a watered-down trial. See the full plan comparison →

Tools like Harvest and Toggl charge $10–$18 per seat per month, which adds up fast for teams. SnapSight's Pro plan is a flat $5/month per user — less than half the cost of most alternatives, with no per-seat scaling surprises. For agencies, the savings compound quickly. See our full feature and pricing comparison →

Yes. Each team member runs SnapSight individually on their own device with their own projects and export access. Multi-user collaboration features (shared dashboards, team rollup reports) are on the roadmap. For agencies tracking multiple clients per person, SnapSight handles the per-person workload well today. See the agency use case →