DoneThat vs Dayflow

The short answer

Dayflow is free, open source, local-first, and good. That is the honest starting point, and any comparison page that buries it is not worth reading.

If what you want is an automatic work journal on a Mac — a timeline of your day with AI summaries, no manual tagging, everything staying on your machine — Dayflow does that, costs nothing, and you can read the source. The case for anything paid has to be made on what Dayflow does not do, not on what it does.

What Dayflow gets right

It is genuinely free. Not freemium, not free-for-one-user. MIT licence, no paid tier. The only cost is whichever model you point it at, and Gemini's free tier is supported, so it can be free end to end.

Everything stays local. Screenshots and data live on your machine. With a local model, nothing leaves at all — which is the strongest privacy position available in this category and stronger than ours.

Bring your own AI. Local models, Gemini, ChatGPT or Claude. You control the provider, the cost and the data path.

It has real traction. 6.1k GitHub stars, and used by people at Meta, Google, DoorDash and Square. This is a maintained project with a community, not a weekend prototype.

Open source means auditable. Not "we describe our capture path" — you can read it.

The standup feature is well judged. One-click standup updates from the timeline is exactly the small, specific thing a daily-standup team actually wants.

Where the two diverge

Not on capture quality. On what happens after capture.

Dayflow is a journal. It records the day, summarises it, and presents it back. The loop closes when you read it.

DoneThat is a coach with a record underneath. Don nudges you onto goals, blocks calendar time for them, flags low-value work while you can still fix it, checks your mood, reminds you of meetings, suggests breaks. The record exists to make that possible; reading it is not the point.

Whether that is worth paying for depends entirely on whether your problem is not knowing or not doing. Journalling solves the first. It does not touch the second.

Screenshots: two different answers

Both products deal with the same uncomfortable fact — understanding screen work means seeing the screen — and they resolve it differently.

Dayflow keeps the screenshots, locally. That is what makes fully offline operation possible. The trade is that an archive of images of your screen now exists on your disk, and it is yours to secure, back up and eventually delete.

DoneThat never stores them. Screenshots are processed in real time and immediately deleted; only the derived structured activity persists. Excluded apps are redacted on your machine before anything is sent, and with your own Gemini or OpenAI-compatible model, raw activity never reaches our servers. The desktop app is open source too.

Neither is strictly safer. One keeps the images under your sole control; the other does not keep them at all. Pick according to which failure you would rather not have.

Platform support

As of 11 August 2026, if you are not on a recent Mac, this decides it.

Where Dayflow is the better choice

• Cost matters at all. It is free, and free is hard to argue with.
• You want everything offline. A local model plus local storage beats any cloud arrangement.
• You are on macOS and expect to stay there.
• You want to read the code. Fully open source, with an active community.
• A journal is what you want. If reading a good summary of your day is the whole requirement, that is what Dayflow is for.
• You do standups. The one-click update is a nice piece of design.

Where DoneThat is the better choice

• You need Windows or Linux today.
• You want to be nudged, not informed. Coaching during the work rather than a summary after it.
• You want goals tracked and something that notices when you drift off them.
• You want AI tools to have your work history. An MCP server and API make it queryable as long-term memory; Dayflow has no equivalent.
• You want a team dimension — summary-level dashboards, relative trends rather than per-person hours, visibility set per summary, optional mood tracking.
• You want peer accountability. Follow others, share summaries, Slack and Discord, accountability groups.
• You want someone to call. Open source means self-supported.

How we compared these

Everything stated about Dayflow comes from its public site, read on 11 August 2026 and linked below. Claims about DoneThat are first-hand: we build it, and we charge for it, which is the bias to read this with.

Dayflow is free and we are not; its offline story is stronger than ours; and its code is fully open where ours is open only for the desktop app. Those are real advantages and we have not tried to talk around them. The narrow claim we would defend is that a journal and a coach are different products, and plenty of people only need the journal.

Questions people ask about Dayflow

Is Dayflow really free?

Yes, entirely — it is open source under the MIT licence with no paid tier. The only cost is whichever AI provider you point it at: Gemini's free tier works, ChatGPT and Claude need their own subscriptions, and a local model costs nothing beyond your hardware.

Does Dayflow work on Windows?

Not yet. As of 11 August 2026 it requires macOS 14 Sonoma or later, with a Windows version announced as coming. DoneThat runs on Windows, macOS and Linux today.

Does Dayflow store screenshots?

Yes, locally on your machine, which is what allows the fully offline local-model workflow. DoneThat takes the opposite approach: screenshots are processed in real time and immediately deleted, with only the derived structured activity kept. Neither is strictly safer — one keeps the images under your control, the other does not keep them at all.

Why pay for DoneThat when Dayflow is free?

Only if you need what it does not do. Dayflow journals your day very well. It has no coaching, no goals, no accountability, no team dimension, no MCP, and no Windows or Linux build. If a local, free, Mac-only work journal is the requirement, Dayflow is the better answer and you should use it.