Guide
You will never remember to press start. Automatic trackers watch for you - here is how the technology works, what it costs, and how to keep it private.
Every timer-based tracker depends on you remembering to start and stop it. Miss either step and the record develops gaps that you have to reconstruct later. Automatic tracking removes that manual step: the computer observes, and you read the results.
Most automatic trackers - Timing, Rize, RescueTime, ActivityWatch - record which app and window title is in front. It is efficient and battle-tested, but it sees labels, not meaning: three hours of "Chrome" could be research, procrastination, or both, and the tracker cannot tell.
Dayflow takes the second approach: it records your screen at one frame every 10 seconds and has AI summarize each 15-minute block into plain language. The result is a visual work journal that captures the task context an app name alone cannot explain.
Dayflow is the only tracker on this page that is fully automatic, AI-powered, open source, and free. Full recordings stay on your Mac. Local AI keeps analysis on-device, while BYO cloud providers and Dayflow Pro can receive selected screenshots for analysis.
| If you want | Use | Price (July 2026) |
|---|---|---|
| A readable journal with local AI available | Dayflow | $0 |
| Invoiceable per-project hours | Timing | From ~$10/mo |
| Focus scores and coaching | Rize | From $9.99/mo annual |
| Distraction blocking | RescueTime | From $7/mo annual |
| Raw activity data on every OS | ActivityWatch | $0 |
Deeper dives: the full 8-app Mac roundup, RescueTime alternatives, Rize alternatives, Timing alternatives, and ActivityWatch vs Dayflow.
Two approaches exist. Metadata trackers (Timing, Rize, RescueTime, ActivityWatch) log which app and window is active and build charts from that. Dayflow instead records the screen itself at one frame every 10 seconds and has AI summarize what was actually happening every 15 minutes - which is how it can tell "researching on YouTube" from "watching cat videos" when both are just "Chrome" to a metadata tracker.
Not meaningfully, if the app is built well. Dayflow uses about 100MB of RAM and under 1% CPU while recording. Choosing local AI for analysis is more demanding (it uses your GPU during summarization), so plug in for long fully-offline sessions; cloud analysis keeps the load near zero.
It depends on where capture and analysis happen. RescueTime, Rize, and Toggl process activity in their clouds. Dayflow keeps full recordings on your Mac and keeps analysis on-device with Ollama or LM Studio. If you use a BYO cloud provider or Dayflow Pro, selected screenshots can be sent to that service for analysis. The open-source Mac client shows local storage and outbound network behavior; the privacy policy covers cloud retention.
In Dayflow, yes - start and stop recording instantly from the menu bar, or automate it with Shortcuts via deeplinks. You can also delete any recordings and data from your local folder at any time.
The commercial trackers run $7 to $29 per month (RescueTime from $7/mo, Rize from $9.99/mo, Timing from ~$10/mo, all billed annually, as of July 2026). Dayflow and ActivityWatch are free and open source. Dayflow has an optional Pro plan ($16/mo annual) that handles the AI for you.
Dayflow tracks automatically and keeps full recordings on your Mac. Choose local AI for on-device analysis.
Download for MacOS macOS 14+ · free & open source · no account required