CodeIt guide
Which coding tool suits my child's age?
What actually works at each age, why the wrong tool at the right age is the usual reason a child quits, and when to move them on. Sourced and dated.
The short version: ages 5 to 7 need pictures rather than words. 8 to 11 is where blocks work and most children start. Around 11 or 12 blocks stop being satisfying and the wrong move here is the usual reason a child quits. For teenagers, adult tools work and are better, and Python is usually the right first text language.
Parents of 8, 9 and 10 year olds ask this most often, and the answer for that group is almost always Scratch first, then a decision at 11 or 12 about whether to move to Python or to the web.
The common mistake is not starting too early. It is leaving a child in a tool they have outgrown, because the subscription is already paid and the tool still technically works.
Everything below is checked against each product's own site, quoted where it matters, on 29 August 2026.
Ages 5 to 7: before reading is fluent
At this age the constraint is not intelligence, it is reading speed. A tool that requires reading each block will exhaust a five-year-old before they build anything.
ScratchJr. Their site states it is for "young children (ages 5-7)". Picture blocks, no text, made for exactly this window. This is the default recommendation.
Kodable. Paid for families: a seven-day trial with no card, then US$24.99 a month, US$119.99 a year, or US$199.99 once, covering up to four children. The free Kickstart plan people mention is on the school plans page, where the paid tiers run US$1,750, US$2,500 and US$3,500 a year per site. Do not assume either price is the other.
Tynker: icon coding for pre-readers is genuinely well built and one of the few paid things at this age worth the money. Family plans are $18/month quarterly, $15/month yearly at $180, or $468 lifetime for up to three children.
Not us. CodeIt asks a child to describe what they want in words. A five-year-old who cannot yet type a sentence will not get far. We are honest about that on our own pricing page and we will be honest about it here.
Ages 8 to 11: the block years
This is the sweet spot for block coding and where most children start and stay for two or three years.
Scratch. "Scratch is free for everyone" (Scratch Foundation). Still the best on-ramp that exists, and the community is a large part of why. If your child is happy here, there is no reason to pay anyone anything.
Google CS First. "totally free of charge — any number of students, all materials, as many lessons as you want." Video-led units built on Scratch. Best if you want structure rather than a blank canvas.
Code.org. Now CodeAI since 2 June 2026, still free, still at the same address. Courses with a path through them.
Where CodeIt starts to fit: the top of this band, around 10 or 11, for a child who wants to make a thing and share it rather than complete a course.
Ages 12 to 15: the gap where children quit
This is the hard part, and it is the reason this site exists.
Blocks start feeling babyish somewhere around 11 or 12. But the adult text-code tools assume motivation a 12-year-old has not built yet, and the drop from dragging a block to a blank file with a blinking cursor is where most children stop coding permanently.
What works: anything where they still make something they care about on day one, but the real code is visible. What fails: a syntax course. They will complete two lessons and never return.
The p5.js Web Editor. p5.js is "a free and open-source JavaScript library", and the editor runs in the browser with nothing to install. Excellent for a 14-year-old who wants to make visual things. No guidance at all, which suits some children and loses others.
Code.org continues to work here, and its later courses are text-based.
Where CodeIt fits, plainly: this band is the whole reason we exist. A learner describes a website, game or quiz, plays it, changes it by moving things and picking colours, and a separate view shows what the finished project is made of and opens the lesson behind each idea in it. Free plan, no card, ten AI-assisted builds a month, lessons and playground free regardless. Paid family plan CA$12 a month in Canadian dollars.
If your 13-year-old is thriving in Scratch, leave them there. This band is about the ones who have stopped.
The one rule that matters more than the age bands
Ages are a starting guess, not a verdict. A confident 9-year-old belongs in the 12-to-15 advice and a hesitant 13-year-old belongs in Scratch, and neither is a problem.
Watch what happens when they get stuck. A child in the right tool gets annoyed and tries again. A child in the wrong tool goes quiet and finds something else to do. That signal is more reliable than any age range on this page, including ours.
Ages 16 and up: use the real thing
At this point the honest advice is to stop using anything designed for children.
freeCodeCamp. "Every aspect of freeCodeCamp is 100% free. The courses, the projects, even the certifications." Project-based, serious, and the right answer for a teenager who has decided they want this.
The p5.js editor, a plain text editor, or whatever a working developer uses. A 16-year-old who wants to build things should be building them with the tools that build things.
Not us, mostly. CodeIt goes to 18 and the lessons are useful for a late beginner, but a motivated 16-year-old is better served by freeCodeCamp and we would rather say so.
| Age | Where most children actually are | What fits |
|---|---|---|
| 5 to 7 | Reading is not fluent yet | Pictures and icons, not words on a screen |
| 8 to 11 | The block years | Scratch and its relatives, where a missing semicolon cannot end the session |
| 12 to 15 | The gap where children quit | Something that produces a visible result fast, and shows the real code behind it |
| 16 and up | Ready for the real thing | The tools adults use, not a version made simpler |
Every row is argued in the section of the same name below, including where we are not the answer.
Common questions
What age should kids start coding? Five is old enough if the tool uses pictures rather than words. The constraint at that age is reading speed, not intelligence. ScratchJr states it is for ages 5 to 7 and is the default recommendation. Most children start between 8 and 11, which is where block coding works best.
Is my child too young to learn to code? Almost certainly not, but the tool matters more than the age. If they cannot read fluently, anything that requires reading each block will exhaust them before they build anything. Use picture and icon tools until reading is comfortable, then blocks.
What is the best coding app for a 6 year old? ScratchJr, which is free and built for exactly this window. Kodable is worth a look but is paid for families, with a seven-day trial rather than a free tier. Tynker's icon coding for pre-readers is genuinely well built and one of the few paid things at this age worth the money. CodeIt is not the answer here: it asks a child to describe what they want in words.
What should an 8 or 10 year old use? Scratch, almost always, and there is no reason to pay anyone if they are happy there. Google CS First adds video-led structure on top of Scratch, and Code.org gives a path through courses. Around 10 or 11 a child who wants to make a thing and share it rather than complete a course starts to outgrow this band.
When should we move from Scratch to Python? Not on a date. Blocks start feeling babyish around 11 or 12, and the wrong move at that point is the usual reason a child quits. Python suits children around 12 and up who asked for it, or who are heading toward school computer science. Below that, setup friction and text-only output cause more quitting than the language is worth.
What works for a beginner teenager? At 16 and up, stop using anything designed for children. freeCodeCamp is free, project-based, serious, and the right answer for a teenager who has decided they want this. The p5.js Web Editor suits visual and generative work. Between 12 and 15 the need is different: something that produces a visible result fast and shows the real code behind it, because a syntax course loses them in two lessons.
How do I know if the tool is wrong for my child? Watch what happens when they get stuck. A child in the right tool gets annoyed and tries again. A child in the wrong tool goes quiet and finds something else to do. That signal is more reliable than any age range on this page, including ours.
Questions parents ask first
What age should kids start coding?
Five is old enough if the tool uses pictures rather than words. The constraint at that age is reading speed, not intelligence. ScratchJr states it is for ages 5 to 7 and is the default recommendation. Most children start between 8 and 11, which is where block coding works best.
Is my child too young to learn to code?
Almost certainly not, but the tool matters more than the age. If they cannot read fluently, anything that requires reading each block will exhaust them before they build anything. Use picture and icon tools until reading is comfortable, then blocks.
What is the best coding app for a 6 year old?
ScratchJr, which is free and built for exactly this window. Kodable is worth a look but is paid for families, with a seven-day trial rather than a free tier. Tynker's icon coding for pre-readers is genuinely well built and one of the few paid things at this age worth the money. CodeIt is not the answer here: it asks a child to describe what they want in words.
What should an 8 or 10 year old use?
Scratch, almost always, and there is no reason to pay anyone if they are happy there. Google CS First adds video-led structure on top of Scratch, and Code.org gives a path through courses. Around 10 or 11 a child who wants to make a thing and share it rather than complete a course starts to outgrow this band.
When should we move from Scratch to Python?
Not on a date. Blocks start feeling babyish around 11 or 12, and the wrong move at that point is the usual reason a child quits. Python suits children around 12 and up who asked for it, or who are heading toward school computer science. Below that, setup friction and text-only output cause more quitting than the language is worth.
What works for a beginner teenager?
At 16 and up, stop using anything designed for children. freeCodeCamp is free, project-based, serious, and the right answer for a teenager who has decided they want this. The p5.js Web Editor suits visual and generative work. Between 12 and 15 the need is different: something that produces a visible result fast and shows the real code behind it, because a syntax course loses them in two lessons.
How do I know if the tool is wrong for my child?
Watch what happens when they get stuck. A child in the right tool gets annoyed and tries again. A child in the wrong tool goes quiet and finds something else to do. That signal is more reliable than any age range on this page, including ours.
What you can do on CodeIt
Last verified 2026-08-29.