The Best Math Products for Kids Ages 8-12

If you're going to try one thing on this list, start here. Khan Academy Math is free, covers every topic from early arithmetic through AP calculus, and adjusts to your child's actual level. The placement diagnostic identifies gaps before your child encounters them.

The mastery system requires understanding before advancement, not just correct answers on a first attempt. Your child can't skip ahead with holes in their knowledge. That builds Persistence. Judgment develops through the feedback loop: attempt, get it wrong, watch the explanation, try a different approach. The child learns to evaluate whether their reasoning makes sense, not just whether their answer matches the key. It's the only free product on this list that I'd recommend without reservation.

The real limitation: Khan Academy is stronger for practice and review than for initial instruction. The video-watch-quiz loop works well for kids who already have some conceptual grounding but need structured repetition. For a child encountering a concept for the first time, the videos alone may not be enough. Khanmigo, the AI tutor add-on ($4/mo or $44/yr), was supposed to fill this gap. It hasn't. Early results have been modest. The recommendation here is the free core platform.

Algebra anxiety usually begins when kids hit symbolic notation. The equals sign, the x, the parentheses. For a child who already believes "I'm not a math person," those symbols confirm the story. DragonBox bypasses the notation entirely. Equation-solving becomes a visual puzzle where the child manipulates objects to isolate a target. They're doing algebra before they know they're doing algebra.

DragonBox is the most rigorously studied game-based math product available. A randomized controlled trial with 1,850 seventh graders found that students using DragonBox showed significantly higher algebra posttest scores than active controls, with an effect size of d = 0.269 (Decker-Woodrow et al.). A follow-up dose-response analysis confirmed that students who solved more problems showed better outcomes (Chan et al.). Engagement depth matters, not just exposure.

This is one of the few math products that builds real mathematical curiosity. Kids want to see what the next puzzle looks like, not just finish the level. They experiment with different approaches because the game rewards exploration, not speed. That's where Curiosity and Judgment develop. For the child who has decided math isn't for them, DragonBox is a backdoor that bypasses the emotional baggage. Note: the pricing model has changed from a one-time purchase to the Kahoot! Kids+ subscription at roughly $3-6 per month.

Every other product on this list requires a screen. Prime Climb requires a kitchen table, two people, and a willingness to lose. Made by Math for Love, it's a board game where players add, subtract, multiply, and divide to race along a number spiral to 101. The color system is the design breakthrough: every prime number has its own color, and composite numbers display the color combinations of their prime factors. A child playing Prime Climb internalizes the structure of multiplication and factoring through play, not memorization.

This is the only physical product on the main list, and it earns its spot by doing something no app on this list does: it creates math conversations between people. When your child rolls a 3 and a 5, they have to decide whether to add, subtract, multiply, or divide, and which pawn to move. The strategic choices require number sense, and the social setting means they're explaining their reasoning out loud. That's where Curiosity and Judgment develop.

At a one-time $25-30 purchase, Prime Climb is the highest-value recommendation on this list. Because the game requires real people in the room, Connection gets built too. For family game night, it's an investment that pays off every time you play it. No subscription. No screen. No login.

Some children don't need a different way to learn math. They need more math, faster, with intellectual challenge that matches. Math Academy is built for that profile. The platform uses a knowledge graph and spaced repetition to map what a student knows, identify gaps, and advance them at their own pace. A motivated 10-year-old working through Math Academy can cover in months what school spreads across years.

The fit question matters here. Math Academy's primary user base skews adult and older teen. The interface is plain: no characters, no rewards, no color. It looks like a college platform, because it was built as one. It rewards discipline and tolerates frustration. For a mature, motivated 10-12 year old who can work independently, it's exceptional. For most 8-year-olds, the demands are too high. OzWrites, a parent-reviewer on Substack, published a detailed balanced review that captures both the potential and the friction. Worth reading before committing.

One design choice that generates strong opinions: Math Academy penalizes incorrect answers by subtracting experience points. A colleague's daughter was doing well through the 5th grade curriculum when she was hit with a large negative XP penalty twice on material she didn't understand. As he put it: "I don't think this UX is helpful for child or parent." Negative feedback in learning products is a live design debate. Some children respond to stakes; others internalize the penalty as confirmation that they're failing. At $49/mo ($588/yr), the commitment is significant. Make sure the fit is right before subscribing.

Kumon is unlike anything else on this list. It's not an app. It's not adaptive. There's no AI, no gamification, no animated characters. It's a worksheet-based, in-center program where children show up, complete assignments, and progress through a structured sequence. Twice a week in person, daily practice at home. The pedagogy hasn't changed meaningfully in decades.

The case for Kumon is the daily habit. For a child who needs structured, consistent practice and whose family can't reliably provide it at home, Kumon delivers that. Persistence is the whole point, and Kumon earns it the old-fashioned way: sheer repetition and discipline. Friends who use it tell me the same story. The first few months are a grind. The child resists. Then something shifts. The routine becomes automatic, and the child stops arguing about doing math because math is just something they do every day. That consistency matters.

The case against is equally strong. Don't expect curiosity or creativity here. The "drill and kill" reputation is not unfair. At $140-200 per month ($1,680-2,400 per year) plus registration fees, families are paying a premium for structure, not innovation. Kumon has not been independently studied at scale, which is itself a notable finding for a program this widespread. Meta-analyses of tutoring interventions show average effects of 0.37 standard deviations (Nickow et al.), far below the inflated claims tutoring companies sometimes make. If your child is already self-motivated, Khan Academy Math provides a similar progression for free.

Most AI math tutors adjust the problems but not the teaching. Your child gets different questions, but the same explanation regardless of where they're confused. Thinkster puts a human tutor in the loop who reviews your child's actual work: not their answers, their thinking. How they approach a problem. Where they hesitate. What they try before giving up. A human notices things no algorithm catches yet.

The price range reflects a real tier system. Plans range from $45-295/mo depending on the level of human coaching, from AI-driven practice at the low end to personalized video sessions and custom learning plans at the top. The core differentiator is the human element, so the entry tier may not deliver the product's real value proposition. Be clear about what you're signing up for. Thinkster launched an AI Coach feature in May 2025 with voice interaction and multilingual support, but the human coaching remains what sets it apart.

The coaching relationship does something apps can't: a real person who knows your child's name and notices when they stop trying. That human accountability builds Connection in a way no app replicates. For parents who want the personalization of a private tutor at a lower price point, Thinkster offers a hybrid that sits between self-serve apps and full tutoring. The tutoring research is encouraging: meta-analyses find tutoring is one of the most effective educational interventions available, with an average effect of 0.37 standard deviations across 96 randomized studies (Nickow et al.), though the original "2 sigma" claim (Bloom) has never been replicated at that magnitude.

An eight-year-old struggling with fractions often needs someone to watch their face, notice confusion, and adjust the explanation in real time. Apps are getting better at this, but a human tutor remains the most responsive form of instruction available. Dojo Tutor delivers that through 1:1 online video sessions, starting at once per week. The format targets families who want a human tutor but can't access or afford in-person programs.

The child this is really for: an eight- or nine-year-old who shuts down in front of an app but opens up when a real person is watching. Persistence develops through the accountability of showing up each week. The tutoring relationship creates a one-on-one bond that strengthens Connection.

The candid limitations: at $30 per session (starting at once weekly, so roughly $120/mo), Dojo Tutor is less expensive than Kumon or Mathnasium but more expensive than any app on this list. The tutoring is broad, not math-specialized. A child who needs deep mathematical challenge beyond grade level will outgrow what a generalist tutor provides. For that profile, Beast Academy for self-directed enrichment or Math Academy for AI-adaptive acceleration are stronger fits. Dojo Tutor's strength is the human connection at a price point that makes tutoring accessible to more families.