REVIEW · 2019-09-03

inbento

A box sold as relaxing, and a 3×3 grid that stops you cold

Steam store ↗

First Impressions

Line up the helpful positive reviews and the same words keep surfacing: cute, chill, cozy, wholesome. And then, usually in the same sentence, a "but." One reviewer called it "a comfort game that secretly lifts weights." The praise and the warning arrive together.

The store page calls it a chill pattern-matching puzzle and adds that zero understanding of cooking is required. What the developer promises is calm. Yet most of the helpful positive reviews open by correcting that promise: I wasn't expecting much, and it turned out far chewier than it looks.

The negative side describes the same gap in inverted words. "Around world 9 the thing that was supposed to be chill just gets annoying." "It's cute, but it isn't what I'd call fun." The two camps are not looking at different games. They are looking at the same gap, and one reads it as a pleasant surprise while the other reads it as a broken promise.

By the numbers, the surprised-and-pleased camp wins by a mile. Steam's overall sits at 96% positive across 1,089 reviews as of 2026-09-15, labelled Overwhelmingly Positive. For a game released the year after Golf Peaks by the same studio, that is a quietly heavy number.

Putting the Mechanics into Words

The board is a 3×3 lunchbox. The goal is to match the picture you are given. But you never place the food directly: you play cards that describe operations. Move. Swap. Copy. Lift now, drop later. Most cards can be rotated, and each level hands you an exact number of them — spend them all, no more and no less.

In the vocabulary of this site, that is verb subtraction taken to an extreme. Nine cells, a handful of verbs. And yet AppUnwrapper wrote in 2019 that "the possibilities seem endless with those three moves." That is the classic symptom of subtraction doing its job.

What makes it work, as I read it, is that the cards carry state. A copy card remembers what it copied; a lift card remembers what it is holding. So there is a second board beside the nine cells: your hand. The combinatorial explosion happens between the two layers, not inside the grid.

That gives it a different failure mode from a line-drawing puzzle like Cosmic Express. There, you get stuck on where the path goes. Here, you get stuck on what order to play in. Change the order of two cards and the outcome is a different picture, so the solution shows up as a permutation problem.

Diagram of inbento's structure: the 3×3 board and the hand of cards that carries stateThe 3×3 board and a hand that carries state — the search space opens between the two layers (diagram)

Teaching

On the learning curve, the review pool splits cleanly in two: the front half is too easy, and the back half spikes. The first complaint is rarer but sharp — AppUnwrapper called the opening four worlds "filler" and wished the difficulty had ramped sooner.

The second complaint is much more common. "Around world 9 it spikes." "World 10 stopped being fun." "The jump around level 11 caught me off guard." When the stuck points cluster on specific worlds like that, it is worth reading as a design fact: the density of combinations steps up right after the last verb has been taught.

What I find interesting is that the teaching itself is almost wordless. A new card arrives on its own, once, with no explanatory text. So the introduction is patient while the escalation is abrupt — a clean demonstration that how well you teach and how fast you escalate are two separate design dials.

There is a relief valve. Levels unlock three at a time, so a wall can be walked around and returned to. It is the same no-dead-end structure as Golf Peaks, and it turns the spike from a wall into a slope with a detour.

The Texture of Difficulty

The negative vocabulary converges on a single claim to a startling degree: brute force, guess-and-check, luck and guesswork, "it felt like trying combinations until one worked." What they are reporting is not that the game is too hard. It is that the texture of the difficulty is unreadable.

I think that deserves to be taken seriously. Nine cells and a few cards sits just outside the size a person can permute entirely in their head. It looks solvable by pure reading and isn't quite. So some players arrive at the answer feeling they hit it rather than saw it. That is a question of search-space size, not of observation resolution.

The positive side describes the same boards as clicking into place. One Japanese review put it memorably: gave up after thirty minutes, came back days later, solved it instantly. That kind of difficulty — the kind that dissolves after you sleep on it — cannot be explained by brute force. The honest reading is that both accounts are true at once.

The absence of a hint system widens the split. Several negative reviews name it directly: being stuck with no optional hint is miserable. Whether to include hints is a design choice, and I argued the general case separately in the design of hint systems. Here I will only note which way it pushes: with no hint, the guesswork feeling gets louder.

The Feel of the Story

What surprised me in the review pool was how much of it is about the story. The narrative is only a handful of wordless illustrations between chapters — a cat mother and her kitten, packing lunches — and yet a large share of the positive reviews stop to mention it. Heartwarming. It turned out to be about becoming a parent.

Several reviews also register the limit: stills between chapters aren't enough, I wanted it to go further. Measured by volume, they're right — there is very little story here. What makes the little there is land, I think, is that the puzzle and the story are about the same object. Packing a lunchbox is the act of looking after someone.

When a story is bolted onto an abstract puzzle, the two usually sit in separate layers. In the terms I used in puzzles with stories and puzzles without, inbento is the rarer case where both share a verb. Arranging the board is arranging it for somebody.

Sources

This piece was written from the user reviews on the Steam store page as of 2026-09-15.

Steam: inbento (Overall 96% positive, 1,089 reviews, Overwhelmingly Positive)

・Read via WebFetch: the top 12 helpful positive reviews, the top 12 negative, the 10 most recent, and 10 Japanese-language reviews

AppUnwrapper's review (2019-09-02)

・No official Steam screenshots are used here. The machine this was written on could not reach Steam's image hosts, so I could not look at the screenshots, and I would not describe images I have not seen. The one diagram is my own.

Closing

From a design standpoint I give it 8.0 — a little harsher than Steam's 96%. The subtraction is beautifully done; getting this much out of nine cells and a few cards is real work. What I dock it for is the padded opening and the guesswork that never quite resolves in the late worlds.

The experience of the difficulty varies by player. The median completion time mentioned across reviews is around four hours, and where the spike lands inside those four hours differs from person to person. Is the easy opening boring or is it a run-up? Is the late fumbling exploration or luck? Those two forks turn the same game into two different games.

It suits people who enjoy permuting a small board in their head. It does not suit people who buy the word "chill" at face value. The store copy tells a small lie — and the players who handed it a 96% rating are the ones who keep kindly correcting it in their reviews.

Where are you with this game?

Press one and the game goes on your shelf. You can see it on your account page.

Reactions (no login)

Anonymous • one of each per visitor per day

Read next

Related reviews