The fastest Sudoku solve with a proper public attribution is Jakub Ondroušek’s 54.44-second solve at the World Sudoku Championship in Bangalore, India, on 15 October 2017, listed by Guinness World Records. Anything faster without a named event and date is entertainment, not a record.
That answer is less tidy than most search results make it sound. Sudoku has world champions, national champions, livestreamed speed solvers and app screenshots, but those are not the same thing. A sudoku world record needs a person, a puzzle, an event, a date, a timing method and independent acceptance of the result. Remove any one of those and you have a claim, not a record.
The catch is that Sudoku difficulty is not a single scale. A puzzle with many clues can still need a nasty step; a puzzle with fewer clues can collapse quickly if the givens point the way. Speed Sudoku records therefore tell you who completed a particular checked grid fastest under particular conditions. They do not prove who would beat everyone on a hard puzzle tomorrow morning.
So who actually holds the fastest Sudoku solve on record?
Jakub Ondroušek of the Czech Republic is the name to know. Guinness World Records attributes the fastest Sudoku solve to him: 54.44 seconds at the World Sudoku Championship in Bangalore, India, on 15 October 2017.
Before the time even matters, that is the cleanest answer: named solver, named event, date. Plenty of speed Sudoku videos show startling solves. Plenty of apps record tiny times. Some may be genuine. They are still weaker evidence than a result accepted at a recognised competition and carried by Guinness.
Ondroušek was not a random contestant who had one lucky minute. He is one of the leading Czech Sudoku and puzzle solvers, and the Czech team has long been strong in World Puzzle Federation events. That matters because a sub-minute Sudoku solve is not normal handwriting with a bit of luck. It is pattern recognition under pressure. The solver sees singles, locked structure and forced placements before most players have finished scanning the first box.
The 54.44-second figure has a boundary. This was not the record for the hardest Sudoku, and it was not a claim that every valid grid can be solved in under a minute. It was a fastest accepted solve of a particular competition Sudoku. For the search query fastest sudoku solve, that distinction is the whole story. The record is real because it is narrow.
What number was it, and how was it counted?
| Claim type | Best documented figure | Attribution | What it really proves |
|---|---|---|---|
| Fastest official single Sudoku solve | 54.44 seconds | Jakub Ondroušek, World Sudoku Championship, Bangalore, India, 15 October 2017; listed by Guinness World Records | A checked competition solve was completed faster than any other publicly attributed Guinness-listed Sudoku time. |
| Earlier widely cited Guinness time | 1 minute 23.93 seconds | Thomas Snyder, Philadelphia Inquirer Sudoku National Championship, Philadelphia, 20 October 2007 | The old benchmark was also event-based, not an app screenshot or a private stopwatch claim. |
| Sudoku championship title | No single solve time | World Sudoku Championship, run through World Puzzle Federation competition structures | A champion wins across rounds and puzzle types; that is not the same as the fastest one-grid solve. |
| Online speed Sudoku claim | No official universal record | No shared authority covers every website, app, puzzle source and input method | Times can be impressive, but generator difficulty, controls and verification vary too much. |
| Hardest fast solve claim | No official universal record | No accepted global standard defines “hardest” for a timed record attempt | Clue count is a poor proxy; the decisive issue is the hardest technique the puzzle actually requires. |
The record is a timed completed grid accepted in a competition setting. It is not counted by clue count, by rating label, or by how dramatic the solve looks afterwards.
Which other speed Sudoku records are worth trusting?
- Thomas Snyder’s 1 minute 23.93 seconds is worth keeping. It has the same useful shape as Ondroušek’s later figure: named solver, named event, named city and date. Snyder made the time at the Philadelphia Inquirer Sudoku National Championship in Philadelphia on 20 October 2007.
- World Sudoku Championship results are trustworthy, but they are not usually fastest-solve records. A sudoku championship rewards consistency across many puzzles, formats and rounds. The winner is not automatically the holder of the fastest single-grid time.
- National championship times can be good evidence when the event publishes enough detail. A named organiser, fixed puzzle set and checked submissions matter. Without those, a time is hard to compare across countries or years.
- Video solves are evidence, not records by themselves. A clear uncut video can show that a solver is exceptionally fast. It still leaves questions about puzzle source, prior exposure, input method and independent checking.
- App leaderboards are the weakest category. They may be fun for speed Sudoku practice, but different apps define difficulty differently and some allow settings that change the task.
How did the record solve happen on the day?
The public version is spare, not cinematic. The public record gives Ondroušek, 54.44 seconds, the World Sudoku Championship, Bangalore, India, and 15 October 2017. It does not give a move-by-move transcript that lets you watch the solve unfold cell by cell.
That missing detail is useful. It stops the story turning into theatre. A record like this happens in a competition room, on a prepared puzzle, under timing and checking rules, surrounded by other solvers who understand exactly how outrageous the pace is. The final number only matters if the completed grid is correct. One wrong digit turns a beautiful minute into nothing.
A sub-minute Sudoku solve is usually decided before the pen has travelled very far. The solver is not debating possibilities in the way a casual player does. They are reading the grid in chunks: rows that force a digit, boxes that have only one possible home, columns that collapse after one placement. The eye gets ahead of the hand. The hand then has to keep up without smudging the result into an error.
It also would not have looked like an expert puzzle grind. Records of this kind need a grid that breaks quickly for a top solver. If a puzzle requires several layers of advanced technique, the clock dies. That does not cheapen the record. Sprinting is a separate skill from climbing. A 54.44-second solve means Ondroušek recognised the route almost immediately and wrote it down correctly at competition speed.
Why do some “records” get repeated and others get ignored?
The repeated records usually have a neat number and a famous wrapper. Guinness is useful because it gives journalists and players something fixed to quote. The ignored records often fail one boring test: nobody outside the solver can verify exactly what happened.
- A named person is not enough. Strong solvers can still make unverifiable claims, and weak solvers can still upload convincing-looking screenshots.
- A date matters. Without it, nobody can tell whether the claim came before or after a better documented one.
- The puzzle source matters. A hand-picked easy grid, a known puzzle and a fresh competition puzzle are different challenges.
- The input method matters. Paper, keyboard, touchscreen and mouse do not take the same time, even when the thinking is identical.
- The checking method matters. A completed grid must be right. Speed without verification is typing.
The biggest trap is clue count. People love saying that a puzzle with fewer givens is harder, but Sudoku does not behave that cleanly. A sparse grid can have a direct path. A fuller grid can hide a step that stops most solvers cold. For records, clue count is an easy number that answers the wrong question.
A proper sudoku world record is not “someone solved a Sudoku very fast”. It is “this named person solved this checked Sudoku at this named event on this date in this time”.
That is why the Ondroušek and Snyder figures survive scrutiny while many faster-looking claims do not. The evidence is attached to the time.
Frequently asked questions
No. An online timer, a paper sheet, and a filmed attempt measure different things. Paper solving includes writing speed and hand movement; online solving includes interface speed, keyboard shortcuts, and input errors. Compare them only when the puzzle, timing start, timing stop, and checking method are the same.
9×9 is the standard grid size; expert describes puzzle difficulty. A very fast record may use an ordinary 9×9 puzzle that is not expert level. Expert Sudoku usually needs harder techniques, such as X-Wings and beyond, so its best times should not be measured against records set on easier grids.
People often think clue count equals difficulty. The correction is that solving path matters more. A puzzle with fewer givens can fall apart through singles, while a puzzle with more givens can require an advanced pattern. A good difficulty grade is based on the hardest technique needed, not the number printed at the start.
For example, a solver may spot that the only possible 7s in two rows sit in the same two columns. That X-Wing removes 7 from those columns elsewhere, and several cells then collapse into singles. In a timed solve, one pattern like that can save more time than ten careful pencil marks.
It depends on the timing rule: whether the clock starts when the puzzle is first visible, when the solver first writes, or on a judge’s signal. The stop point matters too. Some attempts stop on the final entry; others stop after verification. Records are only comparable when those two moments are fixed.
Compared with Mini Sudoku (4×4), standard Sudoku has a much larger search space, more units to track, and more room for advanced patterns. Mini Sudoku (4×4) is excellent for quick logic practice, but its times cannot say much about a 9×9 record. It is a different race.
Trust records with a visible puzzle, visible timer, and independent checking; the exception is an official competition where judges handle those jobs in the room. A bare claim without the grid is weak. A screenshot of a final time is better than nothing, but it does not show whether the puzzle was fresh.
Most people make the same mistake: they check every placement as if they were proving it to a judge. Fast solvers still use logic, but they reduce repeated scanning. They know which houses are nearly full, which digits are constrained, and when a placement will trigger a chain of easy follow-ups.
A mistake is any entry that makes the final grid invalid or forces a correction before completion. In practice, different timers treat mistakes differently. Some allow backspacing during the solve. A formal record needs a correct finished grid, and the timing rules should say whether corrections are allowed before the clock stops.
Practise on Expert Sudoku, then review the point where you slowed down. Look for the first advanced step you missed, not the final time. If X-Wings and harder patterns are still costly, replay similar positions until you recognise them quickly. Speed follows recognition. It rarely follows frantic clicking.
Nobody knows precisely, because app timers, competition sheets, videos, and casual claims are not one clean record book. The best available public benchmark is still the officially reported top time of 1:23.93. Beyond that, there is a large cloud of fast claims with uneven evidence behind them.
The first World Sudoku Championship took place in Lucca, Italy, in 2006. That matters because organised events gave Sudoku records a clearer setting: common puzzles, known competitors, judges, and published results. It did not make every later claim trustworthy, but it raised the standard for what a serious record should show.
The cost of getting it wrong is a false comparison. A solver who finishes an easy grid in record time has done something impressive, but it is not evidence of the fastest expert solve. Difficulty labels need to describe the techniques required. Otherwise a clean speed record turns into a misleading boast.
Final thoughts
If a friend showed me a 54.44-second Sudoku record and then looked disappointed by their own ten-minute solve, I would tell them they are comparing different sports. Speed Sudoku at record level is a sprint on a puzzle that top solvers can read almost at sight. Hard Sudoku is closer to a quiet argument with the grid.
The record is still wonderful. There is something clean about it: one grid, one clock, one correct finish. No long season, no panel of judges deciding style, no bonus for making it look difficult. You either fill the grid correctly before the time stops or you do not.
For ordinary play, the better lesson is not “write faster”. It is “see more before you write”. Fast solvers are not guessing at high speed. They reduce uncertainty early. They notice which digit is nearly placed in a band, which box has been squeezed, which row has become inevitable. The hand moves quickly because the decision has already happened.
So enjoy the record as a sharp little human feat. Then take the pressure off your own clock. A good Sudoku session is allowed to include staring, backtracking in your head, and the small pleasure of finding the one placement that makes the whole grid give way.
-
Sudoku
Can You Cheat at Sudoku?
Sudoku cheating depends on the setting: a solver can break your own puzzle, but only money, rankings, or rules make it serious.
16 August 2026
-
Sudoku
Sudoku X: How Two Diagonals Change the Whole Puzzle
Use the diagonals as extra houses only when they shrink candidates more than the ordinary row, column, or box.
16 August 2026
-
Sudoku
Naked Pairs and Hidden Singles: Sudoku's First Real Techniques
A practical order for using hidden singles and naked pairs, with reasons, traps, and checks you can apply on the next Sudoku grid.
15 August 2026