Hazel: what the free tier covers

The question behind a search for Hazel pricing is rarely just the number. It is whether the free version is a time limited trial or a permanent tier, what disappears when the trial ends, and whether paying once is really paying once. All three have published answers, and the licensing model is different enough from the usual monthly subscription that the arithmetic works out differently too.

The price, and the absence of a subscription

Hazel 6 costs 42 US dollars for a single user license. A family pack covering up to five members of one household costs 65 dollars, and the store states plainly that it is for private households only. Volume pricing exists but is handled by contacting the vendor rather than through the store.

The licensing model is the part worth reading twice. The official FAQ states it directly.

Is Hazel a subscription? No, when you purchase a license, you get free upgrades for that major version. Even if a new major version comes out, your license will still unlock the version you purchased.

Two things follow from that. Point releases inside a major version arrive at no cost, and a license does not expire when the next major version ships. The copy that was paid for keeps running. Nothing stops working because a card expired or a renewal was skipped.

That changes how the cost should be modelled. A monthly tool accumulates cost for as long as it is useful. A one time license creates a cost only when a major version changes, and the release history suggests that happens every few years. Hazel first shipped in 2006, version 5 arrived in November 2020, and version 6 followed after that. Anyone budgeting for this should be thinking in terms of a purchase every several years, not a line item every month.

Three states, and only one of them is time limited

The free part of Hazel is not one thing. The documentation defines three distinct states.

State	When it applies	What is available
Evaluation	First 14 days after install	Every feature except rule export
Demo	Day 15 onward, unregistered	One monitored folder, two active rules, no import or export
Licensed	After registering a license	Everything, permanently

The evaluation period is 14 days and withholds only rule export. Multiple folders, complex nested conditions, every action in the catalogue, all of it runs exactly as it would after purchase. That is long enough to answer the real question, which is whether the filing decisions in a given workflow can be expressed as conditions at all.

When the 14 days are up, nothing is deleted and the app does not stop launching. Hazel drops into demo mode, which watches one folder, keeps two rules active, and turns off import and export. There is no second deadline after that. Demo mode is permanent.

One detail is not documented: which two rules stay active when the switch happens. The safest approach is to decide that in advance rather than discover it on day 15, especially for anyone who spent the evaluation writing a dozen rules across several folders.

What one folder and two rules can actually do

This is the decision point, and it deserves specific thought rather than a shrug. Two rules on one folder is not nothing.

A workable configuration looks like this. Point Hazel at the Downloads folder. The first rule sorts arrivals into subfolders by kind, so images, documents, and archives separate themselves. The second rule moves anything that has not been opened in a set number of weeks to the trash. Between them, those two rules keep the single most chaotic folder on most machines from growing without limit, and that costs nothing.

For a meaningful number of people this is genuinely the whole problem. The mess is concentrated in one location, the sorting logic is shallow, and there is no need to carry the setup to another machine. Paying 42 dollars for capacity that will never be used is not a better outcome than using the demo mode deliberately.

The thing to avoid is drifting into demo mode by accident after building fifteen rules across four folders during the evaluation. Those rules are not deleted, but only two stay active, and working out which two are still running is an unpleasant way to spend an afternoon. Anyone planning to stay free should build for the constraint from the start.

Where the free tier stops being enough

Four situations push past the limits, and three of them arrive sooner than expected.

More than one collecting location is the obvious one. Downloads and Desktop are usually both a mess, and the limit is one folder. Adding an export folder from a camera or a scanner makes it three.

More than two classification branches is the second. Invoices, quotes, screenshots, and everything else cannot be described in two rules without the conditions becoming unreadable.

Processing subfolders is the one people miss. Hazel only matches items at the top level of a watched folder unless a dedicated rule with the Run rules on folder contents action is present. That rule consumes one of the two slots, leaving exactly one rule to do actual work.

Moving the setup between machines is the fourth. Import and export are disabled in demo mode, so a second Mac means rebuilding every rule by hand through the interface.

What to test during the 14 days

The evaluation period exists to produce a decision, and a cautious trial produces no decision at all. Four things are worth establishing before day 15.

Whether the filing logic can be written as conditions at all is the first and most important. Conditions read facts off a file: name, dates, size, text inside the document, Finder tags, Spotlight metadata. A filing rule that depends on knowing which project is currently active cannot be expressed. If that kind of judgement sits at the centre of the mess, no amount of money changes the outcome, and this is the one test that must happen before purchase.

How many folders actually produced work is the second. During evaluation there is no limit, so six folders get added out of enthusiasm. At the end of the period, check which of them ever had a rule fire. If the honest answer is one, demo mode is sufficient.

Whether the operating system is supported is the third, and it is less obvious than it sounds. The Hazel 6 FAQ specifies macOS 13.0 Ventura or later. The installation page in the same user guide says macOS 12 Monterey or later. The two official pages disagree, so on an older system the only reliable check is running the trial on the actual machine.

Rule export is the fourth. It is the single feature withheld during evaluation, and it is also the feature that matters when the setup has to move to a second Mac. Knowing whether that day is coming affects whether demo mode is viable at all.

What macOS already does for nothing

Some of this work does not require a purchase. Smart Folders collect files matching a set of conditions into a single view without moving anything, which covers the case where the only real cost is time spent searching. Tags provide a classification layer independent of the folder hierarchy, so a project can live in folders while its status lives in tags. Shortcuts can bundle a sequence of steps and run them on demand.

The line between these and a rules engine is what triggers them. Built in features run when a person acts. Rules run when a file arrives. If the goal is finding things faster, the free features cover it. If the goal is a folder that stays tidy while nobody is watching, that trigger difference is what the license buys.

Upgrades, and when they cost nothing

Upgrading from any previous version to Hazel 6 costs 20 dollars. The price is the same whether the existing license is single user or family pack, and the license type carries over unchanged after the upgrade.

There is a free path. Anyone who purchased Hazel during 2024 is eligible for a free upgrade to version 6, claimed through a form on the vendor site. If the purchase date is uncertain, the receipt or the date on the stored license file settles it before filing a claim.

Rules and settings carry over between versions. The FAQ says they should all transfer, and then recommends keeping good backups in case a downgrade becomes necessary. That advice earns its place. Rules accumulate over months, and rebuilding a mature set from memory is far more expensive than the license itself.

Discounts, and the store that has them

Discounts do appear. On 4 September 2026, marking the twentieth anniversary of the first release, Noodlesoft announced 35 percent off everything in the store, upgrades included, running through the following Monday night. Anniversary posts have appeared in previous years as well, so a purchase that is not urgent may be worth timing around early September.

There is also a stated way out after paying. The refund policy offers a 30 day guarantee: anyone unsatisfied with the software can request a refund by contacting the vendor within 30 days of purchase, quoting the order number. The same page encourages using the trial period first to confirm the software meets the need, which is a reasonable description of how the evaluation and the guarantee are meant to work together.

That announcement carried a specific warning: use the web store rather than the in-app one, because the discount may not be reflected in the in-app purchase flow. Buying from inside the application during a sale can therefore mean paying full price. Opening the store in a browser first avoids it.

How to decide whether to pay

Two counts settle this, and neither is about the price.

Count repeated file operations for one week. Every time the same kind of file gets moved to the same place, or renamed to the same pattern, that is one tally. This number is what rules remove, and if it is high, 42 dollars is recovered quickly.

Then count something different: how often the active window changes between looking at files, running a command, and working out what should happen to those files. Rules do not reduce this number at all, because investigation is not repetition. A tool built so that files, terminal, and reasoning share one window is aimed at the second count, and the pricing page sets out what that costs by comparison.

What to change first

Run the 14 day evaluation with the rules that would exist in a paid setup, not a cautious subset, then see what survives the drop to one folder and two rules. If that survives comfortably, stay free. If the week of counting shows the real cost is window switching rather than repeated filing, compare what Atriens charges for against the same two numbers before paying for either.