Fluid: what the free tier covers

Fluid costs nothing to use, and the paid tier is five dollars, once. That is the short version, and for most people it is the whole answer. The longer answer is more interesting, because the free tier covers more than people expect, the paid tier covers a narrower slice than people expect, and the price of a one time license is only meaningful once you know how many years of use it buys on the Mac you actually own.

What the free version actually includes

The official about page says the basic version is free without qualifying it. That is not a trial. The core operation, entering a URL and a name and getting a Mac application out of it, sits entirely on the free side of the line.

There is no cap on how many apps can be created. Mail, calendar, a chat tool, three internal admin panels: build all of them and the price does not move. Each generated app lives in the Applications folder and launches on its own, and none of them phones home to check a license.

No account is required either. There is no sign up screen, no subscription, and no upgrade prompt arriving by email later. The download is 6.3 MB, and the front page states that macOS 10.12 or later is required, which is worth checking before any of the pricing discussion matters.

One thing the free tier does not grant is redistribution. The official FAQ asks users not to hand generated apps to friends or coworkers and to point them at the tool instead. Nothing stops it technically, so this is a request rather than a restriction, but a team deployment done in the spirit of that request means each person installs the tool and builds their own copy.

For the common case, free is the whole product. A dedicated Dock icon, a separate window, a place in the application switcher, a custom icon: if the goal is to stop hunting for one tab, the free tier finishes the job.

What five dollars unlocks

Three features sit behind the license, and the official pages name each one along with the menu it lives in.

Pinning an app to the status bar. The app lives in the menu bar at the top of the screen and drops down on click instead of occupying a Dock slot and a window.

Userscripts and userstyles. Custom JavaScript and CSS injected into the page, managed from the Window menu.

Full screen. Entering full screen from the View menu.

Look at that list as a group and the logic is clear. The paid line is drawn around how the window appears and how the page gets modified. Displaying the site and using it normally is not on the paid side at all.

Deciding whether those three are worth five dollars comes down to usage patterns rather than opinion. A tool opened and dismissed forty times a day rewards status bar pinning, because the cost of each retrieval drops. A tool opened at nine in the morning and left there until six does not. Userstyles pay off when there is a specific irritation to remove: a sidebar eating a third of the width, text set too small, a banner that reappears daily. If neither applies, staying on the free tier is not a compromise.

How buying works

The purchase path is slightly unusual and worth describing, because searching for a buy button on the website will not find one. The front page offers three things: a free download, license recovery, and a bug report link. Purchasing happens from inside the application.

Payment goes through PayPal. The license recovery page states this plainly: every app the developer has sold has always been purchased through PayPal, so entering the PayPal email address regenerates every license ever bought and sends them to that address.

That design has a practical consequence for anyone buying today. A lost license key is recoverable as long as the PayPal address is still known and still receiving mail. If that address belongs to an employer or to an account that later gets closed, the recovery path narrows. Filing the purchase email somewhere durable at the time of purchase takes ten seconds and removes the problem.

There is also no invoicing apparatus around this. The tool is sold by an individual developer, not by a company with a procurement portal, so the paper trail for an expense claim is a PayPal transaction record and an email. For a five dollar line item that is usually fine, but it is a different process from buying software from a vendor that issues invoices.

Dividing the price by the years it has left

A one time price is only half of a cost calculation. The other half is the denominator, and in this case the denominator is published.

The shipping build supports Intel architecture only, which means Apple silicon Macs run it through translation. Apple has set a date on that.

Support for Rosetta will end in a future version of macOS, so check with the app's developer for an updated version. Source: support.apple.com

The same Apple support page is specific about the timing: Rosetta remains available through macOS 27, the next major release, and from macOS 28 it survives only for certain older, unmaintained games that rely on Intel based frameworks.

On an Apple silicon Mac, then, the window for using this tool has a visible edge. Five dollars is small enough that the arithmetic barely matters, and nobody is going to regret the money. What deserves a decision is the other cost: the time spent configuring whitelists, tuning icons and getting sign in flows working inside a container that has a known end date. That investment does not transfer to whatever comes next.

On an Intel Mac none of this applies, and the same five dollars buys an indefinite runway. Same price, two different products, depending on the hardware.

What the money does not buy

A license is not a subscription to future versions, and in this case that distinction has teeth. The update feed the app checks is still online and still answers, and the newest entry in it carries a build date of October 2018. Paying unlocks three features in the build that exists. It does not put anyone in a queue for a build that does not.

Support is the other thing to calibrate. The developer publishes bug reporting guidelines and asks for reports by email, with a template covering the hardware, the OS version and the exact app version. That is a considered process rather than an abandoned mailbox, and it is also a single person answering mail about several products. Expectations should be set accordingly: this is not a support desk with a response time attached to it.

Nor does the license cover the wrapped services themselves. A paid account at the service being wrapped is still a paid account at that service, and nothing about the container changes what the service charges or what it allows. That sounds obvious written down, and it still catches people who assume a wrapper adds something the web version withheld.

If this is for a team

Multiply before deciding. Five dollars per person is trivial at five people and still trivial at fifty, so cost is never the blocker at this scale. What scales badly is the configuration: every person builds their own app, enters their own sign in domains, and repeats that work on every machine they use, because settings live inside each generated app rather than in an account.

The practical consequence is that team rollouts are priced in hours rather than dollars, regardless of which tool is chosen. The variable worth comparing between tools is how many of those hours disappear because the services are already described for you.

The zero cost routes and what they cost instead

Several routes reach the same destination, and comparing only the price column hides the differences.

Route	Price	Architecture	Interface language
Browser install to Dock	Free	Follows the browser	Follows the browser
Fluid, free tier	Free	Intel only	English
Fluid, licensed	$5 one time	Intel only	English
Current site to app tools	Free to one time purchase	Varies	Varies

Chrome and Edge will both install a page as a windowed app at no cost, and that path stays current because the browser stays current. What it does not give is much control after the fact: the window is a browser artifact, and it inherits the browser's updates, profiles and quirks.

The cost people forget to count is setup time. A service without a ready made preset needs its sign in domains discovered and entered by hand, and for a service that bounces through two hostnames during authentication that can absorb half an hour. Measured against an hourly rate, five dollars is smaller than that half hour. Sorting the table by price alone makes exactly the wrong trade visible.

Current tools in this category tend to divide free from paid by app count rather than by feature. The Pricing page for this one shows the pattern: three apps at no cost with nothing held back, and a one time purchase beyond that. The advantage of that structure is that the paid features can be evaluated before any money changes hands.

Three questions that settle it

How many times a day is the service opened and dismissed? Under ten, status bar pinning is decoration. Over ten, it changes the feel of the workday.

Is there a specific visual annoyance to remove? If a sidebar, a font size or a recurring banner has a name, userstyles are the feature that addresses it. If the only complaint was "it takes too long to find", that was already solved by giving it a window.

How long will this service still be in use? Something replaced at the end of the fiscal year does not justify configuration effort in any tool. Something that will be open every working day for years justifies checking whether the tool itself is still being maintained before anything else.

There is a fourth question worth asking if the first three came out close. Does the service in question already have a preset in whichever tool is being considered? A preset removes the configuration work entirely, and configuration work is the only expensive part of this decision. If the answer is yes, the cheapest route is usually also the fastest one.

Three answers of no means the free tier is the correct stopping point. One yes makes the five dollars worth spending, if the hardware makes that purchase last.

What to change first

Check whether the Mac is Intel or Apple silicon, because that single fact decides whether this is a purchase with a deadline attached. Then list the services opened most often and check them against the preset list of whatever tool is in play, such as the Supported services page, since presets are what remove the setup time that costs more than any license here.