Friday, October 30, 2026 Independent reviews No paid placements How we evaluate
Smart Toy Choice

Independent reviews · Reader-supported

Smart toys, honestly reviewed

Independent reviews of smart toys, coding robots and STEM kits for kids

How we evaluate Browse sections

Latest reviews

Browse by section

01

Coding Robots

Two completely different products get sold under the same name. A programmable robot is a machine a child writes instructions for. A companion robot is a machine that talks back. They share a shape and a price bracket, and they teach opposite things — one rewards precision, the other rewards conversation. That confusion is expensive, because the buying advice is not transferable. A programmable robot is worth judging on its programming interface and the quality of its motors. A companion robot is worth judging on what it says and what it does with the audio. Buy the wrong one and the child gets a toy that does something impressive once. This section is about the programmable side: what a child actually writes, at what age each programming style becomes usable, and which parts break first. The robot most people ask us about, <a href="/moxie-robot/">Moxie</a>, sits on the companion side and is covered separately — it is worth reading precisely because it shows what happens when the talking stops. We have not run these robots in a classroom. What we have is published specifications, the actual programming environments where they are documented, and the failure patterns in owner reports after six months of use.

1 review 02

STEM Kits

"STEM kit" is a marketing category, not a technical one. Nothing regulates the term. A box of plastic gears with a leaflet about photosynthesis and a $300 robotics set with a curriculum both get called STEM, and the two have almost nothing in common beyond the word. What separates them is not the price and it is not the branding. It is whether the thing gives a child something to do after the instructions run out. A kit that ends when the model is finished is a craft project. A kit that keeps producing new problems is a tool. That distinction decides whether a purchase gets used for three weeks or three years, and it is invisible in every product photograph. This section currently leads with <a href="/building-toys-like-lego/">building toys and what actually fits together</a>, because construction is the entry point for most families and the place where the most money gets wasted. From there, the same questions apply to electronics kits, chemistry sets and coding hardware — the failure modes are near-identical. We do not run kits ourselves. Everything here comes from published parts lists, replacement-part availability, and the pattern in reviews from families who have owned the thing long enough for the novelty to wear off.

1 review 03

Kids' Tech

Tech for kids fails in a predictable way: the spec sheet is honest and the product is not. A camera lists its megapixels and says nothing about a shutter button small hands cannot reach. Headphones advertise a volume limit and leave out that the limit only applies to one of the two ways a child will actually use them. A tablet quotes a screen size and stays quiet about which apps it will refuse to run in two years. This section covers the three devices parents ask us about most: <a href="/kids-camera/">cameras</a>, <a href="/kids-headphones/">headphones</a>, and <a href="/kids-tablet/">tablets</a>. They look like different categories and they are not. All three sit at the same decision point — the age where a child wants a real device rather than a toy version of one, and the point where a cheap device becomes more expensive than a good one because you buy it twice. Our approach is to judge these products on the thing the box does not measure: what happens in month four. Does the camera still get picked up? Do the headphones still fit the same head? Is the tablet still receiving security updates, or has the manufacturer quietly stopped? Every review in this section states the month a price was checked, names the specific failure we would expect, and says who should not buy the thing. Our full method is on the <a href="/how-we-test/">how we evaluate</a> page.

3 reviews 04

AI Toys

The word "AI" on a toy box currently means at least three different things, and the difference between them is the difference between a product that will last five years and one that will stop working when a company runs out of money. At the bottom: a fixed set of recorded responses triggered by keywords. This is not new and it is not AI, but it is what most toys under $50 are actually doing. In the middle: a microphone that streams audio to a server, where a large language model generates a reply. That is genuinely a language model, and the toy is a terminal for it. At the top, and still rare: a model small enough to run on the device itself, which keeps working when the network does not. Almost every buying decision in this category comes down to which of the three you are holding, and almost no product page will tell you. This section starts with the <a href="/moxie-robot/">Moxie shutdown</a>, because it is the case study that makes the distinction concrete — a genuinely sophisticated robot that became inert when its maker closed, not because anything broke but because there was nothing left to connect to. We have not run these toys with a child. What follows comes from published specifications, teardown reporting where it exists, privacy policies, and the documented behaviour of the models involved. Where we are inferring, we say so.

1 review 05

Safety & Privacy

A smart toy passes the same physical safety tests as any other toy. That part is working. What no label covers is the part that makes it smart: a microphone that stays on, a camera that uploads to an account, an app that asks a five-year-old for a birthday, and a server in a country you will never visit. The question parents actually ask — <em>is this safe for my kid</em> — is three questions stacked on top of each other. Will it choke, burn or cut? Does it listen, record or transmit? And what happens to all of that when the company that built it stops existing? Each one has a different regulator, a different label, and a different way of failing. This section starts with <a href="/are-smart-toys-safe/">what the safety marks actually certify</a> and stops exactly where the marks stop. The Moxie shutdown — a working robot turned into a brick overnight — is covered in <a href="/moxie-robot/">its own piece</a>, because it is the clearest example of the risk nobody insures against: the toy is fine, the company is gone. We do not run a physical testing lab and we do not pretend to. What we do is read the standards, read the regulatory filings, and read the reports from people who own the thing. Where a claim cannot be checked, we say that instead of repeating the manufacturer's wording.

1 review 06

Gift Guides

Most gift guides for children are written in October and wrong by January. They rank products by how they look in a photo, how impressive the box is, and how recently the brand pitched them — none of which predicts whether the thing is still being used in February. The child opens it, the adults photograph it, and by the end of the week it is on the shelf behind the other things that were opened last year. This section takes a different test: <strong>what is still out of the box in month two.</strong> That single question removes most of what gets recommended elsewhere — the novelty electronics, the kits with two hundred pieces and no reason to build them twice, the movie tie-ins that stop being interesting when the film leaves cinemas. Our gift guides are organised by age and by what kind of child you are buying for, because a nine-year-old who builds and a nine-year-old who plays sport need different answers to the same question. Each pick states a real price with the month it was checked, says what the ongoing cost is after the box is open, and names the child who should not get it. If you are buying for a specific age and do not want to read a long article, start with <a href="/gifts-for-10-year-old-boys/">gifts for 10 year old boys</a>. It explains the reasoning, and the same logic applies to the other ages.

6 reviews

Why you can use these verdicts

Every article is checked against records you can open — recall notices, app update histories, dated prices and owner reports. Where we did not have the product in hand, the article says so.

01

Checked, not assumed

Specifications come from manuals and certification filings, not the marketing page. Every figure points at a source.

02

Still works as a service?

Last app update, whether the company is trading, whether the subscription is still sold. A dead backend turns a toy into a brick.

03

Bad news first

If something is not worth the money, that is in the first paragraph. We publish “not recommended” and “insufficient evidence”.

Read the full methodology

About this site

Smart Toy Choice reviews the toys that need charging, pairing, and a parent who reads the manual. Coding robots, STEM kits, kids’ cameras and headphones, AI companions — the category where a $150 purchase can either open a door or sit in a drawer by week two.

We are not a general toy site. We do not cover plush, puzzles, board games, or anything that runs on imagination alone. There are good sites for that. We cover the narrow, expensive, confusing slice where technology meets childhood — and where most buying advice online is written by people who never took the thing out of the box.

What we cover

  • Coding robots — screen-free programmable toys, app-driven robots, and the age at which each one stops being fun.
  • STEM kits — circuits, mechanics, science sets, and subscription boxes that survive past the first weekend.
  • Kids’ tech — cameras, headphones, smartwatches, tablets, drones, and microscopes built for small hands.
  • AI toys — companion robots and talking plush. The newest, least documented, and most over-marketed category we cover.
  • Safety and privacy — what connected toys actually collect, and the questions to ask before you buy one.

How we decide

Every article answers three questions a product page never will: at what age a child can use this without an adult sitting beside them, how much of the work falls on the parent, and whether the thing still works as a service six months after you buy it.

We do not publish star ratings, because a number implies we measured something. Instead we research each product against public records — manuals, certification filings, recall notices, app update histories, dated prices and owner reports — and end with one of four labels: Recommended, Recommended with conditions, Not recommended, or Insufficient evidence. Where a product is genuinely in our hands, the article says so at the top.

Read the full methodology — the six checks every article passes, and what each label means.

Where to start

If you are buying your first coding toy, start with age fit rather than features: the most capable robot on the market is a bad purchase for a five-year-old. If you are choosing between two specific models, our comparison pieces exist for exactly that decision. And if the purchase is for someone else’s child, our gift guides filter by age and interest instead of price.

We publish as research finishes, not on a promotional calendar. Tell us what you are trying to decide — reader requests decide a real share of what we cover next.