Palworld Review (2026): Is It Still Worth Playing?

Updated September 16, 2026. Player counts, version numbers, prices and hardware requirements below were read that day from Steam, SteamDB, Pocketpair's own announcements and Xbox Wire, all linked at the end. Where I give an hours figure for a stage of the game, that is my own structural estimate and I say so; Pocketpair does not publish completion times.

Palworld is no longer the game that broke Steam in January 2024, and it is no longer in Early Access either. It left Early Access on July 10, 2026, which means the question "is it still worth playing" now has a different shape than it did a year ago. The useful way to answer it is not with a score out of ten. It is to ask what each block of hours actually hands you, and at what point the game stops handing you anything new.

Where the hours go

The stage boundaries below are my reading of the game's structure, not an official figure. The content each stage unlocks is taken from Pocketpair's published changelogs.

Stage	What opens up	What has already started repeating
Hours 0 to 5	Capture loop, first base, first few of the 287 Pals	Nothing yet
Hours 5 to 20	Base automation, Pal work suitabilities, tech tree, first bosses	Manual resource gathering
Hours 20 to 50	Sunreach, The World Tree story content, Awakening and Mutation systems	Combat encounters, base layout problems
Hours 50 plus	Breeding optimisation, dedicated-server play up to 32 people, September's two new sub-missions	Most of the moment-to-moment game

Hours 0 to 5: the capture loop does the selling

The opening is still the most persuasive part of the game, and it is persuasive for a reason that has nothing to do with the joke the marketing leaned on in 2024. Capturing a creature in Palworld is a survival-crafting resource action, not a collection action. The Pal you throw a sphere at becomes a worker, and within the first hour you are already thinking about which one you want for its labour rather than its design.

Five hours is enough to see the whole pitch and roughly enough to decide. If the loop of "see thing, catch thing, put thing to work" has not landed by hour five, the next forty-five hours are a larger version of the same loop and will not convert you. That is a genuinely useful thing to know before buying rather than after.

Hours 5 to 20: it quietly becomes a factory game

This is the stretch where the game changes genre without telling you. Pals have work suitabilities, bases have production chains, and by hour fifteen the interesting decisions are about throughput and layout rather than exploration. Anyone who came in expecting a monster-collecting adventure is by now playing something closer to a light Factorio with cute staff, and that is where the majority of the positive word of mouth actually comes from.

It is also where the first real annoyance lives. Base Pals make scheduling decisions you did not ask for, which is precisely why Pocketpair is still shipping fixes for it: the September 6 update, version 1.0.4, adjusts Pal and base-Pal skill priority and charging-skill behaviour. Two and a half years after launch and two months after 1.0, base Pal behaviour is still an active patch target. Treat that as a signal about the shape of the game rather than a bug report.

Hours 20 to 50: the 1.0 content, and the first real wall

The July 1.0 release is the reason this review needed rewriting. According to the official changelog it added a floating sky island called Sunreach, story content under the name The World Tree, 72 new Pals bringing the roster to 287, and two new progression systems, Awakening and Mutation, alongside a full stat and ability rebalance of the existing roster. That is a substantial amount of mid-to-late-game content dropped onto a game whose late game was previously its weakest part.

It also moves the wall rather than removing it. Somewhere in this range the game stops giving you new verbs and starts giving you bigger numbers, and whether that is a wall or a comfortable plateau depends entirely on whether the factory layer grabbed you at hour fifteen. If it did, Awakening and Mutation are the systems you will spend the next hundred hours inside. If it did not, this is where you stop.

Past 50 hours: what actually repeats

Being honest about the repetition is the whole point of an "is it still worth it" review. Past fifty hours the exploration is done, the tech tree is done, and what remains is breeding optimisation, base rebuilding, and multiplayer. Multiplayer is worth its own line: co-op is capped at four players and dedicated servers at 32, per the official store page. Those two numbers describe two very different games. Four-player co-op is the same game with friends. A 32-player dedicated server is a different thing entirely, closer to a persistent shared world, and it is the most reliable way to get another hundred hours out of the game.

The September 6 update added two sub-missions, Beyond Dimensions, which starts at Fisherman's Point in Mount Obsidian, and From the Far Side of the Moon, which unlocks after it. Two sub-missions is a modest drop. If your question is "will there be enough new content to justify coming back every couple of months," the honest answer from the current cadence is: not on its own, but the 1.0 content is genuinely large enough to justify one return.

Two mechanics that work differently than older videos show

This matters if you are watching 2024 footage to decide. Pocketpair stated publicly in May 2025 that two gameplay changes were made as a direct result of the Nintendo patent litigation filed in September 2024. Patch 0.3.11, on November 30, 2024, removed the ability to summon a Pal by throwing the sphere, replacing it with a static summon beside the player. Patch 0.5.5 changed gliding so that it requires an equipped glider item rather than being performed via a Pal directly, with Pals now providing passive glide bonuses instead. Pocketpair described these as compromises made to avoid further disruption to development and said explicitly that the changes were a result of the ongoing litigation.

Neither change is fatal and both are small in play. But if you watch a launch-window video and then buy the game expecting to throw a sphere and have a Pal come out swinging, you will notice on day one. As of this writing I could find no official statement from either Pocketpair or Nintendo dated later than May 8, 2025, and no official announcement of a ruling or settlement, so the case appears to still be open. Anything you read about hearing dates and likely outcomes is commentary, not a company or court statement.

Reading the player numbers without being fooled by either side

There are three numbers in circulation and they measure three different things.

Figure	Value	What it actually tells you
Steam all-time concurrent peak	2,101,867 on January 27, 2024	The launch phenomenon. Says nothing about today.
Pocketpair, after 1.0	Over 850,000 concurrent on Steam, announced July 13, 2026	The 1.0 spike. Official, and two months stale.
SteamDB, checked September 16, 2026	56,471 playing, 24-hour peak 90,486	The actual steady state you would be joining.

Pocketpair also announced 40 million total players on July 8, 2026. Note the wording: players, not copies sold. The company has not published a copies-sold figure, and the sales estimates that circulated in July came from third-party analysts rather than Pocketpair, so I am not quoting them. There is also a small inconsistency in Pocketpair's own numbers worth flagging: the 1.0 changelog says 287 Pals, while a September 1 Steam post describes the roster as "over 280+." That is rounding rather than contradiction, but it is the kind of thing worth noticing before you treat any single figure as precise.

Ninety thousand concurrent players two months after a full release is a healthy, populated game. It is also about four percent of the launch peak. Both of those sentences are true, and which one gets quoted at you usually tells you what the person quoting is trying to sell.

What the recommended spec is really protecting

Steam lists a minimum of Windows 10 64-bit, an Intel Core i5-9400F, 16 GB of RAM, a GeForce GTX 1660 and 40 GB on an SSD. The recommended tier is Windows 11, an i5-12400 or Ryzen 5 5600X, 32 GB of RAM and an RTX 3060 Ti or RX 6700 XT. The store page adds the detail that actually matters: performance varies with the number of active Pals.

That is not a graphics warning, it is a simulation warning. The reason the recommended RAM figure is 32 GB rather than 16 is the base-automation layer described above, which is the part of the game you will be in most after hour twenty. If you are sitting at the minimum spec, expect the opening five hours to run fine and the hour-thirty factory base to be where your machine struggles. Budget your upgrade decision around that stage, not around the tutorial.

Who should not buy this

Skip it if you want a finished narrative experience; the story content is a frame around systems, not the reason to play. Skip it if base-building and production chains bore you, because that is the middle sixty percent of the runtime. Skip it if you are buying primarily for the creature roster, since 287 Pals sounds like variety but functionally resolves into a few dozen work roles. And skip it if you are hoping the litigation gets resolved in a way that restores the original throw-to-summon mechanic, because there is no official indication that is coming.

My call

At the Steam list price of 29.99 US dollars, the arithmetic is straightforward: if you get through hour twenty, the cost per hour is trivial, and the 1.0 content means there is a genuine late game to get to now, which was not true at this time last year. If you have Game Pass Ultimate, Game Pass Premium or PC Game Pass, Palworld 1.0 has been included since July 10, 2026, which removes the decision entirely; play five hours and find out.

Where I would actually spend the money instead is on the multiplayer. The single-player 1.0 experience is good. The 32-player dedicated server version of this game is the one people are still playing in September, and it is the reason the concurrent count is 90,000 rather than 9,000. If you have three or more people who will commit, buy it. If you are playing alone and the factory layer does not appeal, the honest recommendation is to look at something else. Our Minecraft review for 2026 covers the obvious alternative, and the current hottest games roundup covers the rest.

Tip: The 32-player server play described above is the part of Palworld that keeps people past hour fifty, and it is voice-chat driven, so a gaming headset matters more here than in most survival games, and a gaming controller suits the long base-building sessions better than a mouse. (These are Amazon Associate links - we may earn a small commission on qualifying purchases.)

Where these figures came from

• Palworld Steam charts, SteamDB (all-time peak and current concurrents)
• Palworld on Steam (price, system requirements, player caps)
• Palworld v1.0 official release changelog, July 9, 2026
• Palworld v1.0.4 patch notes, September 2026
• Pocketpair: Palworld reaches 40 million total players, July 8, 2026
• Pocketpair: over 850,000 concurrent players on Steam, July 13, 2026
• Pocketpair: regarding the lawsuit, changes to Palworld and the future, May 8, 2025
• Nintendo press release on the patent infringement lawsuit, September 19, 2024
• Xbox Wire: Game Pass July 2026 wave 1

Concurrent-player figures are a snapshot taken on September 16, 2026 and move daily. No retailer discount, sale price or star rating is quoted anywhere in this article; the only price given is the publisher's US list price on Steam.