AMP vs CK-12

CK-12 is free, made by a nonprofit, and reaches past AMP into calculus, science and nearly 50 languages2,3. Much of it is led by the learner: books to read and a chat tutor to ask3,5. AMP costs $599 a year for the first child and makes those choices itself: it finds the level, picks every problem, and explains every answer with worked steps.

Try it free

Why AMP

AMP decides what comes next

CK-12's math comes as digital textbooks, adaptive practice and a chat tutor2. Common Sense Media notes that younger children may need a parent's help to find their material3. In CK-12's own study of its tutor, teachers said students struggled with prompting it5. AMP keeps a level in each of 14 strands of math and picks every problem from the learner's answers: new work, a prerequisite, or review.

The research: Practice at the learner’s level

Every answer is checked exactly and explained

Kid reviewers on Common Sense Media's CK-12 page, as the page summarizes them, reported unclear quiz questions, answers marked right or wrong incorrectly, and a lack of clear explanations3. No AI model runs while a learner works in AMP. Every problem comes from tested code, and 185 independent checkers re-derive the answers. After each answer the learner sees right or wrong, the correct answer and the worked steps.

The research: Feedback after every answer

Help gives the first step, never the answer

CK-12's tutor, Flexi, is a chat box a student types questions into1. In CK-12's own study, teachers worried about students leaning on it too much and about plagiarism5. That study did not test Flexi. In a field experiment with nearly 1,000 high school math students in Turkey, unrestricted GPT-4 raised practice scores by 48%, and once access was removed those students scored 17% lower on exams than students who never had AI; a tutor version with learning safeguards largely mitigated that harm6. AMP builds the safeguard into its rules: a hint is one step that never contains the answer, and an answer after a hint does not count toward mastery.

The research: One-step hints · The research: Mastery, and repair of gaps

It starts at counting

CK-12's math FlexBooks begin at grade 12. Common Sense Media finds most of the site suited to older children and teens3. AMP starts with counting in kindergarten, finds a young learner's level inside the first 24 problems with no separate test, and needs no adult to plan the next lesson.

The research: Practice at the learner’s level

Learned means a rule anyone can read

CK-12's efficacy page reports that two in three students improve their skill level, without defining a skill level or naming a comparison group4. In AMP a skill is Strong after four of the last five answers are right without help. Two misses in a row send practice back to what the skill builds on. The rule sits in AMP's public source code.

The research: Mastery, and repair of gaps

Side by side

AMPCK-12
Price$599 a year or $59 a month for the first child, then $399 or $39 for each more. Two parents practice free. Full refund within 30 days.Free. CK-12 is a nonprofit and calls Flexi a free tutor for every student2. Common Sense Media lists the price as free3. We found no paid tier.
Who can buyFamilies, tutors and schools, directly. Free to start without a card: 100 answers and 5 workbooks. Arizona ESA, Arkansas EFA and New Hampshire EFA funds pay for it.Anyone. Students, families and teachers sign up free. Teachers assign work with a class code1,2.
RangeCounting through precalculus in one program: 336 skills on one map.FlexBooks from grade 1 through calculus, plus college algebra and college precalculus. No kindergarten math listed2.
Finding the starting pointNo separate test. Placement takes the first 24 problems of practice. After that, every answer adjusts a level in each of 14 strands of math.CK-12 says its AI can judge in 2 to 4 questions whether a student is at, below or above grade on a concept. No accuracy figure given4.
TeachingA one-step hint and a written lesson with a worked example for every problem type. Worked steps after every answer. Narrated lesson films.Digital textbooks with text, video, audio and quizzes3, plus Flexi, an AI chat tutor1,2.
What decides the next problemAMP chooses every problem from the learner’s answers: new work just within reach, a prerequisite, or review.Learners and teachers pick the book and lesson3. An adaptive practice product exists; we found no published rule for how it picks items2.
MasteryFour of the last five answers right without help. An answer after a hint doesn’t count. Two misses in a row send practice back to what the skill builds on.No published mastery rule found. Its efficacy page reports skill-level gains without defining a skill level4.
What the adult has to doNothing to plan, teach or grade. AMP places the learner, chooses each problem, checks it and schedules review.Common Sense Media says younger children may need a parent's help to find their material3.
ReportsThe level in every strand with its likely range, skills, time, and the actual problems behind each number. For teachers, who needs them today.A teacher-facing Insights feature4. Links with Google Classroom, Canvas and Schoology2.
EvidenceA published rationale (ESSA Tier 4) with the research behind each part, public source code, and outcome studies from 2027, each registered before it starts.No math entry in Evidence for ESSA and no WWC intervention report7,8. Its own studies report teacher opinions, small tests with no scores shown, and statistics with no stated method4,5.

What CK-12 is

From CK-12 Foundation

CK-12 comes from the CK-12 Foundation, a nonprofit that gives its materials away2. Its math is a set of FlexBooks, digital textbooks from grade 1 through algebra, geometry, trigonometry, precalculus and calculus, with college algebra and college precalculus above them2. The same site covers science, English, social studies and more2.

A FlexBook lesson mixes text, video, audio and quizzes3, and the site also lists interactive science simulations and PLIX interactive items2. Beside the books sits Flexi, an AI tutor a student asks in a chat box, on the web, in a Google Play app or on WhatsApp1,2. Teachers assign work with a class code and connect CK-12 to Google Classroom, Canvas or Schoology1,2.

CK-12 also offers adaptive practice2. How it chooses items is not described on the pages we read, so this page does not guess at it.

What CK-12 does well

The evidence

CK-12

CK-12's efficacy page lists its studies4. The fullest is a six-week study of Flexi in spring 2025, run by CK-12 with Leanlab Education: ten middle and high school math and science teachers in five states, studied through surveys, diaries and focus groups5. It reports what teachers saw, with no test scores. Teachers reported more engagement, confidence and curiosity. They also reported that students struggled with prompting, that lower-readiness learners had trouble with comprehension, and worries about overreliance and plagiarism5.

The other entries are small studies, such as a blind rating of Flexi against Khanmigo by 25 students and a five-day pre-test and post-test with 76 students in grades 7 and 8 who used Flexi to study; the page lists no results for the rating and no scores or effect sizes for the test4. It also gives platform statistics with no stated method. One says two of three students improve their skill level, with no sample size, date, comparison group or definition4. Without a comparison group, figures like these cannot show what the feature caused. The student page calls Flexi the "World’s Most Powerful AI Tutor"1.

We found no CK-12 math study with a comparison group and published test results, on its site or in either national clearinghouse as of November 2026. Evidence for ESSA has no math entry for CK-12; its one CK-12 page, filed under science, reads "No studies met inclusion requirements"7. The What Works Clearinghouse has no intervention report on CK-128.

AMP

AMP's rules follow findings that have held up across many studies. Across 40 studies of computer-based feedback, feedback that explains averaged an effect of 0.49 (the studies mix ages and subjects), the correct answer alone 0.32, and right-or-wrong only 0.059. AMP shows the worked steps after every answer. Across 55 studies, from elementary school to college, worked examples improved math performance by g = 0.4810. Every AMP problem type has a lesson with one. We expect that structure, applied at every problem, to show in outcomes. That is our bet, and we will test it in public.

AMP is new, and its own outcome studies have not run yet. They begin in 2027: growth on a national test against hours of use, then comparison groups, then randomized trials. Each will be registered before it starts and published whatever it finds. Until then, the research behind each part of AMP is laid out in our report, and the code is public.

The research behind AMP

Choose CK-12 if

Choose AMP if

Questions

Is CK-12 really free?

Yes, as far as we found. CK-12 is a nonprofit and calls Flexi a free tutor for every student2, and Common Sense Media lists the site's price as free3. We found no paid tier.

Why pay for AMP when CK-12 is free?

AMP costs $599 a year or $59 a month for the first child, and $399 or $39 for each more, with two parents practicing free. The price buys a program that places the learner, picks every problem, explains every answer, schedules review and prints workbooks with answer keys, from counting through precalculus. The first 100 answers are free, and there is a full refund within 30 days.

Does AMP have an AI tutor like Flexi?

No. Flexi is CK-12's AI chat tutor1. No AI model runs while a learner works in AMP: every problem, hint and explanation comes from tested code, and 185 independent checkers re-derive the answers.

Can we use CK-12 and AMP together?

Yes. A learner can read a FlexBook or ask Flexi about a new idea and practice in AMP. CK-12 also covers calculus and science, which AMP does not2.

Is AMP proven to work better than CK-12?

Not yet; the evidence section above says where AMP's studies stand. We found no CK-12 math study with a comparison group and published test results either4,5.

Sources

  1. CK-12 for students (CK-12 Foundation)
  2. CK-12 mission, products and FlexBooks (CK-12 Foundation)
  3. CK-12 website review (Common Sense Media)
  4. Efficacy studies (CK-12 Foundation)
  5. Flexi, CK-12's AI student tutor: Spring 2025 study (CK-12 Foundation with Leanlab Education)
  6. Bastani et al. (2025), Generative AI without guardrails can harm learning (PNAS)
  7. CK-12 Series (Evidence for ESSA)
  8. Intervention reports (What Works Clearinghouse)
  9. Van der Kleij et al. (2015), Effects of feedback in a computer-based learning environment
  10. Barbieri et al. (2023), A meta-analysis of the worked examples effect on mathematics performance

Checked November 2026.

CK-12 is a trademark of its owner. AMP is not affiliated with it. Found an error? Write to skylar@corpora.inc.

More comparisons

All comparisons · AMP · Research