Khan Academy vs a Live Tutor: What Works Better in 2026

Khan Academy's own research shows it works. That same research shows only a small fraction of students actually use it enough to get those results. That gap, not whether the platform is "good," is the real question behind Khan Academy vs a live tutor.

The short answer: Khan Academy can produce real, measurable learning gains, but only for students who use it consistently, roughly 30 minutes a week or more, which most kids don't do on their own. A live tutor costs money largely because a person is there to make sure that consistency actually happens, not because the instruction itself is dramatically more advanced.

Here's what the research on each actually shows, and how to tell which one, or which combination, fits your child.

Yes, for the students who use it enough. In a November 2024 efficacy analysis, students who used Khan Academy for 30 or more minutes a week, about 18 hours over a school year, showed roughly 20 percent greater than expected learning gains on MAP Growth assessments across grades 3 to 8, an effect size of 0.36. Even lower usage, 9 to 18 hours a year, still produced smaller but real gains of about 7 percent.

The catch is who actually reaches that level. Only about 9 percent of the roughly 350,000 students in that research hit the recommended 30-minutes-a-week threshold. Nationally, the typical student spends closer to 14 minutes a day when they do log in, but that's spread unevenly, with a small share of highly engaged users pulling the average up while most students use it far less consistently. The platform genuinely works. Getting a specific child to use it enough, on their own, without anyone checking in, is the part that usually doesn't happen.

This is where a famous number gets misused. A now-classic 1984 paper by Benjamin Bloom claimed one-to-one tutoring produced a two standard deviation improvement over conventional classroom instruction, informally known as the "two sigma problem," an effect so large it would be almost unheard of in education research today.

More recent analysis has walked that number back substantially. A 2011 meta-analytic review found the real average effect of human tutoring closer to 0.79, less than half of Bloom's original claim, and traced much of the original gap to the fact that tutored students in Bloom's study were held to a stricter 90 percent mastery standard before moving on, not to tutoring itself being inherently magic. In other words, a meaningful share of tutoring's advantage comes from enforcing mastery before advancing, the exact mechanism that's optional, and frequently skipped, in a self-directed tool like Khan Academy.

Factor	Khan Academy	Live Tutor
Cost	Free	Typically $35-$115 per session for structured 1:1 programs
Format	Self-paced videos and practice problems, no live instruction	Live, real-time instruction with a person
Accountability	None built in; depends entirely on the student's own follow-through	Built in; a scheduled session and a tutor who notices when something isn't landing
What the research supports	Real gains, but only at 30+ minutes a week of consistent use	Gains driven largely by enforced mastery and consistent attendance, not format alone
Personalization	Adaptive practice based on right or wrong answers	Adaptive based on a real-time read of confusion, hesitation, and reasoning, not just correctness
Best fit for	Self-motivated students who mainly need more practice reps, or a free supplement to any curriculum	Students who need someone to notice a specific gap, keep pace, and hold the line on mastery

Which should you choose? Khan Academy is a genuinely research-backed free resource, but that same research shows most kids won't use it enough on their own to see the gains it's capable of. If your child is self-motivated and mainly needs more practice reps, Khan Academy alone can work well. If your child needs someone to notice a specific gap, keep them accountable for showing up, and enforce mastery before moving on, a live tutor is doing a job Khan Academy was never designed to do.

1. Time how many minutes your child spends on Khan Academy in a typical week without you reminding them. Under 30 minutes puts them below the threshold tied to real gains, so self-directed use likely isn't enough on its own yet.

2. Ask your child to explain one skill they practiced this week without looking anything up. If they can't, they were clicking through problems rather than mastering them, regardless of which platform they used.

3. Check whether the same type of problem keeps tripping them up across multiple sessions. A tool that doesn't flag or address a recurring gap is a sign your child needs a person watching for it, not just more repetitions.

4. If you're leaning on Khan Academy as a free supplement, put a specific weekly time on the calendar rather than "whenever they have time," since that's the exact follow-through gap the research points to.

5. If your child has tried consistent, timed Khan Academy practice for a few weeks and the same gap persists, that's less about which platform and more about needing someone to diagnose exactly where the breakdown is.

Is Khan Academy actually effective?
Yes, according to Khan Academy's own November 2024 efficacy research, students who used it 30 or more minutes weekly saw about 20 percent greater than expected learning gains. The caveat is that only about 9 percent of students in that research actually reached that level of use.

Is Khan Academy enough to replace a tutor?
For a self-motivated student who mainly wants extra practice, it can be enough on its own. For a student with a specific skill gap, or attendance and motivation issues, tutoring research suggests a live person adds accountability that a self-directed platform doesn't provide.

How much better is a live tutor than self-paced practice, really?
Modern research puts the real-world effect of tutoring closer to a 0.79 effect size, smaller than the often-cited "two sigma" claim from 1984, and a meaningful share of that gap comes from tutoring enforcing mastery, not from the live format by itself.

Can I combine Khan Academy and a live tutor?
Yes. Many families use Khan Academy for extra reps between tutoring sessions, since a tutor can point to specific skills worth practicing, which raises the odds a student actually puts in enough time to see gains.

How many minutes a week does my child need on Khan Academy to see results?
Khan Academy's own research ties meaningful gains to about 30 minutes a week, roughly 18 hours over a school year, with smaller but still measurable gains starting around 9 hours a year.

Cosmo isn't trying to replace free tools like Khan Academy, and a tutor will often point a student straight back to resources like the ones in Cosmo's roundup of free math practice websites for extra reps between sessions. What Cosmo adds is the part self-directed practice consistently struggles with: a Personalized Learning Plan that diagnoses exactly where a specific skill gap is, a live teacher who notices in real time when something isn't landing, and a monthly report so progress is visible instead of assumed.

That combination tends to matter most for elementary and middle schoolers who haven't yet built the independence to self-monitor their own practice, which is why Cosmo has separate options for elementary math tutoring and middle school math tutoring for grades 6-8, each built around a student's actual skill level rather than their grade alone.

Try a free trial class with Cosmo and see whether your child needs more practice, or a person making sure the practice actually sticks.