How to get faster at the Rubik's Cube
Cubeathon guide · Reviewed October 2026
Getting faster isn't one trick. It is a handful of skills (a short cross, smooth F2L, quick recognition, few pauses) improved one at a time, with a timer telling you what actually worked.
This guide goes through them roughly in the order they pay off, from your first sub-minute solves to the milestones beyond. If you can't solve the cube yet, start with the beginner course.
1. Know your current level
Time a batch of 12 solves with proper scrambles and look at your Ao5 and Ao12, not your best single. Averages tell you your real level; a single fast solve might just have been an easy scramble.
Write down where you are now. Every change you make later is measured against it.
2. Turn smoothly, not frantically
Fast cubers look fast because they rarely stop, not because each turn is violent. Aim for an even rhythm you can keep for the whole solve. Rushing leads to lockups (a layer catching mid-turn) and to mistakes that cost far more than the time you saved.
A cube that turns freely helps; a stiff or very loose one makes smooth turning harder. Learn the common finger tricksfor R, U and their primes early, so you don't have to unlearn whole-hand turning later.
3. Plan the cross in inspection
Any cross can be solved in eight moves or fewer, yet beginners often take far more. Use the 15 seconds of inspection to plan the whole cross before you start, then solve it on the bottom so you can see the F2L pieces on top.
Practise it on its own: scramble, inspect, solve only the cross, check how many moves it took. Doing the cross with your eyes closed after planning it is a classic way to check the plan was complete.
4. Learn F2L
F2L(first two layers) replaces the beginner method's separate corner and middle-layer steps: pair a corner with its edge, then insert the pair. It is usually the biggest single speed-up after the beginner method, and the largest part of every CFOP solve.
Learn it intuitively first. Understand the few basic inserts and why they work, then solve pairs by reasoning rather than memorising all 41 cases.
5. Improve lookahead
Lookahead means finding your next pieces while you are still solving the current ones. Without it, every F2L pair starts with a pause while you search.
The standard way to train it is the slow solve: turn deliberately slowly, so slowly that you never have to stop. While one pair goes in, your eyes are already looking for the next. It feels unnatural at first; it is also one of the most reliable ways to break a plateau.
6. Cut the pauses
In most solves, more time goes into stopping than into turning. Watch for where you stop: searching for the cross pieces, looking for the next pair, working out which OLL or PLL case you have.
Each kind of pause has its own fix. Planning more in inspection helps the first, lookahead the second, and recognition drills the third. Recognising a last-layer case instantly is a skill in itself; practise it separately from turning.
7. Learn OLL and PLL
Start with 2-look: 10 algorithms for OLL and 6 for PLL, solving the last layer in four looks. Then learn full PLL, all 21 cases, a few at a time. Full OLL (57 cases) can come later and take as long as it takes.
Learning an algorithm properly
Watch it, set the case up yourself, do it slowly until your hands know it, and drill recognition until you can name the case without thinking. An algorithm you have to recall mid-solve isn't saving time yet.
8. Practise deliberately
Doing hundreds of solves on autopilot makes you good at your current habits. Better practice has a focus: one session for the cross, one for slow solves, one for drilling a few new algorithms, then timed solves to see whether it carried over.
Short, regular sessions beat occasional marathons. Learn a few new algorithms at a time and use them in solves before adding more.
9. Measure progress over weeks, not solves
Single times bounce around; your Ao12 and Ao100 show the trend. Expect plateaus: weeks with no change are normal, and are usually a sign it is time to change what you practise.
10. Milestones: sub-60 to sub-10
Cubers mark progress by sub-Xmilestones: averaging under a minute, under 30 seconds, under 20. Here is what usually makes the difference at each one. It reflects common community advice, not a strict rule; everyone's path is a little different.
Sub-60
Solve with the beginner method from memory, without looking anything up, and read notation fluently. Most of the time here goes into remembering what comes next; once the steps are automatic, sub-60 usually follows.
Sub-30
Plan the cross in inspection and solve it on the bottom, then learn 2-look OLL and PLL. Swapping the beginner last layer for 16 algorithms removes a lot of repeated sequences.
Sub-20
Intuitive F2L done without long pauses, full PLL, and basic finger tricks. This is where lookahead starts to matter more than learning new algorithms.
Sub-15
Efficient F2L (fewer moves per pair, fewer rotations), more of full OLL, and smooth turning. Many cubers also start planning the first F2L pair during inspection.
Sub-10
Full OLL and PLL, advanced F2L choices, very little pausing, and a lot of consistent practice. Few cubers get here, and the ones who do usually took years.
Not sure where you stand? See what counts as a good time.
11. Try a competition
You don't need to be fast to enter a WCA competition. Most events welcome first-timers, and solving under real conditions (a judge, inspection, a competition timer) is a different and useful kind of practice.