Didon icon

Didon

Didon is an AI time tracker for Mac that automatically logs work by analyzing screenshots on your device. It helps users replace manual timers with local AI-based work logs, project breakdowns, and daily summaries.

Didon

Overview

Didon is an AI time tracker for Mac that automatically logs work by analyzing screenshots on your device. It is designed to reduce manual time entry and turn your activity into a daily work log, project breakdown, and summary report.

The product asks you to define projects and categories first, then monitors your screen in the background every 30 seconds. Its local AI model reads window titles, visible text, and on-screen context to infer what you are doing, classify the work, and save logs locally for later review. The site also positions the app as a way to understand work habits, not just count hours.

Features

Automatic background tracking

Didon runs quietly in the background and captures a screenshot every 30 seconds so it can infer activity without manual timers.

Context-based categorization

You can define projects and activity categories up front, and the AI uses that context to classify work and build statistics based on your setup.

On-device AI analysis

The app analyzes screenshots locally on your Mac with a local LLM (Qwen-3-VL:2b), which keeps screen data on-device rather than sending it to the cloud.

Detailed activity detection

Didon identifies windows, file names, apps, and visible task context to generate a structured work journal and daily work log.

Local CSV work log exports

The app saves exports locally, including per-day CSV work logs and a daily summary report file, making it easier to review or reuse the data later.

Auto-launch and auto-pause

Didon starts with your system and pauses when your Mac sleeps, so tracked time is meant to follow your active work session automatically.

Use cases

  • Automatic daily time tracking

    Useful for freelancers and independent workers who want a running record of how much time went to coding, marketing, research, or other task categories without starting a timer each time they switch tasks.

  • Daily work log review

    Helpful when you want to review where your time actually went after the day is over, using project totals, category breakdowns, and local CSV reports as a work history.

  • Project-based categorization

    Fits users who want their current projects classified against a defined list, so work tied to a client or internal project can be separated from uncategorized activity.

  • Local-first privacy workflow

    Appropriate for privacy-conscious users who want AI-assisted analysis without sending screenshots to a cloud service, since the site says processing stays on the Mac.

  • Work habit analysis

    Useful for people trying to spot work habits and productivity patterns, since the product is positioned to reveal focus time, activity distribution, and workday trends.

Pros and Cons

Pros

  • No manual start/stop timers are required once monitoring is active.
  • Screen analysis happens locally on the device, which keeps screenshots and logs off the cloud.
  • Project and category context help tailor the classification to how you actually organize your work.
  • Didon generates both detailed daily logs and summary reports that can be reviewed after the fact.
  • The app is set up to auto-launch and auto-pause, which reduces missed time when sessions start or stop unexpectedly.

Cons

  • The app is currently limited to Mac on Apple Silicon (M1+), so it is not broadly cross-platform yet.
  • Accuracy depends on the quality of your project names and file names; the site recommends using descriptive labels instead of generic ones.
  • The current workflow appears centered on local CSV exports and built-in reports, while broader integrations are described as upcoming rather than available now.

FAQ

How does Didon know what I’m working on?

Didon captures a screenshot of your screen every 30 seconds while monitoring is active. Each screenshot is analyzed locally on your Mac, using visible window titles, text, and context to infer what you are working on.

How does Didon recognize my projects?

You add your active projects in the Configuration tab as comma-separated names. Didon compares those names with what appears on screen, including file names, project names, and browser tab titles, and labels work that does not match as "Other".

Can I export my work logs?

Didon saves local exports in ~/Documents/Didon/. The source mentions detailed work logs in WorkLogs/YYYY-MM-DD.csv and summary reports in Daily-Reports.csv, and says you can review them for productivity analysis.

What platforms does Didon support?

Didon is currently available for Mac on Apple Silicon (M1+). The site also says a Windows version is planned and that users can join a waitlist for early access.

Is my data private?

The site says all screenshot analysis and data processing happen locally on your Mac, and that screenshots and activity logs never leave your machine. It also notes the local model loads briefly during analysis and uses around 2 GB of RAM.

Quick Facts

Category
AI time tracker
Platform
Mac (Apple Silicon, M1+) only at present
Core workflow
Automatic screen-based work logging with local AI analysis
Output
CSV work logs and daily summary reports
Data processing
On-device, local LLM analysis
Website
didon.app