PomoFit runs on whatever interval you give it. Two settings cover almost everyone: 50/10 and 90/20. Open the timer, set the block, and when it ends you get a break with something in it rather than a five-minute gap you spend on your phone. Free tier, no account.
The arithmetic that kills the 25-minute block
Every block of real work carries a fixed cost at the front. Opening the right files. Rebuilding the model in your head. Remembering which of the four things you tried yesterday was the one that actually failed. Call it spin-up.
Spin-up does not get cheaper because the block is short. That is the whole problem.
Suppose yours runs ten minutes, which is modest for anything architectural. Inside a 25-minute block that leaves fifteen usable minutes, and you have handed 40% of the session to overhead. Put the same ten minutes at the front of a 90-minute block and the overhead drops to 11%. Fifteen usable minutes against eighty. Same person, same task, same warm-up.
Nobody has measured your number and no study sets it at ten. But you already know roughly what it costs you to get back into a hard problem, and you know the countdown fires at about the moment it stops hurting.
Worth saying plainly: 25 is not a scientific quantity either. It was chosen, not measured, by Francesco Cirillo in the late 1980s, and the object he chose it with was a kitchen timer shaped like a tomato. No optimal interval is waiting to be discovered. There is only the length that fits the work in front of you, and for work that needs context loaded, 25 is short.
50/10 and 90/20, and how to choose
50/10. The sane default. Long enough that spin-up stops dominating, short enough that you will actually finish the block instead of abandoning it at minute 62. Start here for a week before you touch anything else.
90/20. For the sessions you plan the day around. New architecture, a migration, a hard proof, a first draft of something long. One or two of these is a full day of that kind of work, and pretending otherwise is how people end up disappointed with every focus system they try.
Shorter than 50. Fine, but be honest about what for. Review queues, triage, email, admin. Those genuinely are 25-minute tasks and the shorter interval suits them.
If you are already running the classic interval and it works, keep it. This page is for the people it does not work for. The full argument about who those people are sits in why the Pomodoro technique stops working.
A twenty-minute break is a different problem from a five-minute one
Most of the break research is about micro-breaks, meaning ten minutes or less. The best pooled estimate comes from Albulescu and colleagues in PLOS ONE in 2022: 22 samples, 2,335 people. Vigour rose and fatigue fell, both at around a third of a standard deviation (d=0.36 and d=0.35). Task performance came out at d=0.16 and missed significance. Narrowed to cognitively demanding tasks, the estimate was d=−0.09, which is to say nothing at all.
One detail in that paper matters more at 90/20 than anywhere else. The breaks that ran longer were the ones associated with better performance afterwards. That is not proof that twenty minutes off makes you better at your job. It means whatever faint signal exists is sitting at the long end of the range rather than the short end.
Which leaves you with a practical question nobody asks about 5-minute breaks: what do you actually do with twenty minutes?
Leave the room. Not the desk. The room.
Do something with a load in it. Twenty minutes is enough for a genuine set, not a token one. The full body strength plan is built at roughly this length.
Eat, if it is that time. A break long enough to eat in is a break you will not resent.
Do not open the ticket queue. A break spent on a different screen is not a break, it is a context switch with worse ergonomics.
PomoFit fills that window for you, which is the only reason it exists as a separate product. When the block ends, the screen shows a movement sequence and a clock. You can skip it. There is no streak to protect and nothing lights up when you comply.
Two blocks is a real day
The honest ceiling is lower than the productivity internet suggests. Two 90-minute blocks, or three 50-minute ones, and the rest of the day goes to the work that does not need a runway. Meetings, review, replies, the small pile.
Planning four deep blocks and hitting two feels like failure. Planning two and hitting two does not. Same output, different Friday.
When your calendar is already chopped up
Long intervals assume you own your day. Plenty of people do not.
Measure the real gaps before you pick an interval. If your longest uninterrupted stretch is 55 minutes, 90/20 is a fantasy and 50/10 is the plan.
Put the block at the front of the gap, not the middle. Spin-up before a meeting is spin-up wasted.
Treat a 12-minute gap as a break, not a block. That is the slot the exercise goes in, and it is the one people waste.
Stack meetings deliberately if you can. Three back to back leaves one real block. Three scattered leaves none.
The same logic applied to a home-office day is in Pomodoro for remote workers, and the version written for people shipping code is at Pomodoro timer for programmers.
What you should not expect from it
Not a higher IQ. Movement in a break lifts energy and mood dependably: pooling 158 studies, Reed and Ones put the rise in activated positive affect at d=0.47, still measurable thirty minutes later, and low-intensity work produced it just as consistently as hard effort. Cognition is a weaker story. Reviewing 113 studies in 2024, Garrett and colleagues put the acute effect of exercise on thinking at g=0.13, and reported strong signs that the literature behind it is skewed by selective publication.
So: energy, yes. Mood, yes. A body that still works in twenty years, plausibly. Sharper thinking, no, and anyone selling you that is selling.
The most reassuring finding is the flat one. A 2026 paper in the British Journal of Sports Medicine recruited 19,342 US adults, of whom 11,484 finished a fortnight of five-minute movement breaks. Engagement with work rose 4 to 7%, self-rated performance 1 to 3%. There was no control group and everything was self-reported, so the percentages are soft. What survives is the sign on them. Stopping to move did not take anything away.
Setting it up
Set the interval on the timer and start. Nothing to install, no account, and the free tier is a working product rather than a trial.
Pro costs $4.99 a month. It buys scheduling that adapts to you, exercises you define yourself, ambient audio from YouTube, proper statistics and leaderboards. If you would rather be handed a break sequence than assemble one, the ready-made plans are the shortcut. Weekly brain fog killer suits long-block days better than the cardio ones do.
Frequently Asked Questions
What is a deep work timer?
A deep work timer is a focus timer set to a long interval, typically 50 or 90 minutes, rather than the classic 25. The longer block suits tasks with a high start-up cost, like writing, coding or design, where a short countdown ends the session shortly after you have finished loading context.
Is 90 minutes the ideal length for a focus block?
No length is ideal, and the popular 90-minute figure is often justified by ultradian rhythm claims that remain contested. Ninety minutes is useful because it is long relative to the time most people need to get going, not because a study identified it. Pick the length your calendar can actually protect.
Should I take a 20-minute break after 90 minutes of work?
A proportional break is a reasonable default, and the 2022 micro-break meta-analysis found longer breaks did more for subsequent performance than very short ones. Twenty minutes is also long enough to leave the room, move properly and eat, which five minutes is not. Shorten it if the day is tight.
Do exercise breaks improve deep work?
They improve how you feel, not how you think. Pooled evidence shows movement reliably raises energy and lifts mood, while studies measuring cognition after exercise or sitting breaks have found little or nothing. The practical case is that breaks restore energy without costing output.
Can I use PomoFit for free?
Yes. No account, no card, and the free version carries the timer itself, intervals you can set to any length, and guided movement in every break. Pro costs $4.99 monthly for adaptive scheduling, your own exercises, ambient audio, statistics and leaderboards. Two long blocks a day rarely needs it.
What if I get interrupted mid-block?
Stop the block and restart it rather than pretending the interruption did not happen. A 90-minute block with a ten-minute conversation in the middle is two short blocks and two spin-ups. Knowing that is more useful than protecting a number on a screen.