How long should a Steam demo be?

Design 30 to 60 minutes of content for a linear game, and one or two full runs for a replayable game, then judge the result by median playtime, not by your design. That is our suggestion. It rests on three facts read today. Valve sets no length. Most players stop well before the end. And the demos of games that sold more held their players longer.

In Chris Zukowski's 2022 benchmark of 130 demos, most developers said they designed their demo to run about 30 minutes, yet the median demo had a median playtime of 14 minutes. His warning: "the average player will not play your game for the entire time you designed it for." So if you want players to spend 20 minutes, you need more than 20 minutes of content.

Mark Brown, who made the Word Play demo, gives a wider target: "You should aim for somewhere between 30 and 90 minutes." He checks it with the "median time played" stat in Steamworks, and reports that Word Play sits near 1 hour.

The numbers behind the answer, each from a dated source read 19 November 2026.
WhatNumberSource
Length most developers designed forAbout 30 minutesHow To Market A Game, 2022, 130 demos
Median of the median demo playtime14 minutes (10 at the 30th percentile, 22 at the 70th)How To Market A Game, 2022
Median playtime, demos of top revenue tier games65 minutes (Diamond), 38 minutes (Gold)How To Market A Game, 2022, 62 released games
Suggested target by one developer30 to 90 minutesGame Maker's Toolkit, 2025
Length rule for Next FestNoneValve, Steam Next Fest documentation

Sources: How To Market A Game (26 October 2022, read 19 November 2026); Game Maker's Toolkit (9 June 2025, read 19 November 2026).

What does Valve say about demo length?

Valve gives a balance, not a number. The Demos page says demos "can range in size and scope" and asks you to give players enough to get excited "without giving away so much that they feel like they've experienced everything the game has to offer." It also says players use the demo "to make a purchase decision," so quality counts more than length.

For Steam Next Fest, Valve says "there are no requirements regarding exclusivity, newness, or length of demo." The demo must be public by the time the fest begins, and Valve asks you to submit it for review at least 3 weeks before, or 5 weeks before to be live for the Press Preview. Valve also encourages you to leave the demo up after the fest, because "Many players have expressed disappointment when demos disappear immediately."

One more rule shapes the timing. Once the demo is playable for the first time, you can send a notification to players who wishlisted the full game, but only once and within 14 days of that first release. Have the length right before you launch the demo, not after.

Valve's words on demo length and content, quoted from Steamworks pages read 19 November 2026.
TopicWhat Valve says
What a demo is"generally are a small playable portion of your game that show some of the core mechanics" (Demos)
How long it can be"can range in size and scope" (Demos)
The balance to strike"You need to balance giving the customer enough content to get them excited, without giving away so much that they feel like they've experienced everything the game has to offer." (Demos)
Quality"Make sure that what you put out there is high quality and measures up to the real experience." (Demos)
Next Fest length rule"there are no requirements regarding exclusivity, newness, or length of demo" (Steam Next Fest)
Next Fest deadline"Submit your demo for review no later than 3 weeks prior to the Fest" (Steam Next Fest)

Sources: Steamworks Demos and Steam Next Fest documentation, read 19 November 2026.

Does a longer demo get more wishlists?

Longer playtime goes with more sales, but a longer demo does not cause it. Zukowski estimated revenue for 62 released games from their review counts and sorted them into four tiers. The median demo playtime rose with each tier, from 7 minutes for Bronze to 65 minutes for Diamond. The Diamond tier also held one co-op party game with a very low median, which he reads as players trying it alone and coming back with friends.

His point is about holding players, not about adding minutes: "That is the biggest issue, not whether it is linear or not but whether people want to play through it to the end." A demo that bores players for an hour is worse than one that thrills them for 20 minutes.

Wishlists from demo players are a separate number. In the June 2026 Next Fest data that Zukowski collected, a median of 19.3% of demo players wishlisted the game, with 15.9% at the 30th percentile and 23.2% at the 70th. None of these sources links that rate to demo length, so we do not either.

Median demo playtime by later revenue tier, How To Market A Game 2022 benchmark (read 19 November 2026). Revenue is the author's estimate from review counts.
Revenue tierAverage demo playtimeMedian demo playtime
Bronze8 minutes7 minutes
Silver19 minutes18 minutes
Gold43 minutes38 minutes
Diamond56 minutes65 minutes
Median demo playtime by revenue tier (minutes)Bronze: 7Silver: 18Gold: 38Diamond: 65

Sources: How To Market A Game (26 October 2022; 30 June 2026), both read 19 November 2026.

How long were demos that did well?

Much longer than the median, and the replayable ones longest. GameDiscoverCo reported a median demo playtime of 1 hour and 5 minutes for Cloverpit, a roguelite, when it had nearly 150k Steam wishlists. Cairn, a single player survival climber, had a 44 minute median and more than 440,000 demo downloads on Steam. Guildrun, a PvE autobattler roguelike, reached a median of 3 hours 55 minutes about 10 days after its demo came out, with a quarter of players over 10 hours.

These are single games that GameDiscoverCo chose because they did well, so they show what is possible, not what is typical. Each of them is a game players want to replay or keep climbing in. A linear story game should not expect a 4 hour median.

Brown's Word Play shows how to cap a replayable demo. He limited content, not time: the demo has about 25 perks and one mode, and he made sure "the demo files only contained 25 perks and 1 round."

Reported demo playtimes of single games that did well. Not a benchmark. GameDiscoverCo, read 19 November 2026.
Game and genreMedian demo playtimeReported at the time
Cairn, survival climber44 minutesMore than 440,000 Steam demo downloads (1 July 2025)
Cloverpit, roguelite1 hour 5 minutesNearly 150k Steam wishlists (29 April 2025)
Guildrun, PvE autobattler roguelike3 hours 55 minutesAbout 10 days after the demo, over 275,000 players (28 July 2026)

Sources: GameDiscoverCo (29 April 2025; 1 July 2025; 28 July 2026); Game Maker's Toolkit (9 June 2025); all read 19 November 2026.

Does the right length depend on the genre?

Yes, mostly on whether the game is linear or replayable. That is the only split the 2022 benchmark tested. Linear demos had a median of 14 minutes, with 9 minutes at the 30th percentile and 20 at the 70th. Replayable demos had a median of 18.5 minutes, with 10 and 28. Zukowski notes that sandbox, roguelike and procedural games can push the numbers up, and that the top replayable games ran nearly 20 minutes longer than the top linear one.

No source we read today gives a reliable length per Steam tag, so the genre rows in the decision table below are our suggestion. The genre guides in this series go deeper: a horror demo is limited by how long a player can stand being scared, a cozy or farming demo should show one full loop, and a roguelike or deckbuilder demo should make players start another run.

Median demo playtime, linear and replayable demos, How To Market A Game 2022 (read 19 November 2026).
Kind of demo30th percentileMedian70th percentile
Linear9 minutes14 minutes20 minutes
Replayable10 minutes18.5 minutes28 minutes

Sources: How To Market A Game (26 October 2022, read 19 November 2026); our genre demo guides.

What should you cut from a demo?

Cut whatever stands between a new player and the core loop. Zukowski played many of the top 50 demos of the June 2024 Next Fest and found that many had no tutorial, or only a popup that listed the controls. His advice: "You have to assume you have players' attention for minutes." He also sees median playtime drop a bit after a festival, and guesses that players give many games "a quick 1-5 minute shot."

Then cut what belongs to the full game. Valve's balance applies here: keep enough to excite, and keep the rest for launch. A good rule of ours is to keep the best 30 minutes and cap the content, not the clock.

  • Long intros, logos and cutscenes before the first input. Let players skip them, or move them after the first fight or puzzle.
  • Tutorials that stop play. Teach in the level, or show the controls once.
  • Slow early grind. Start the demo a little later in the game, with some upgrades already unlocked.
  • Late game systems. Show them on the end screen or in the trailer instead.
  • Anything that crashes or confuses testers. Valve's build review only checks "that the game has the features you've promised and it runs," so the QA is yours.

Sources: How To Market A Game (24 June 2024, read 19 November 2026); Steamworks Demos and Steam Next Fest documentation, read 19 November 2026; our practice.

Where should the demo end?

On a high point, followed at once by a call to wishlist. End after the first big reveal, a boss, a full run or a finished day, while the player still wants more. Then show an end screen that names the full game, says what the full game adds, gives the release window and has a Wishlist on Steam button.

Steam helps here. Valve says the demo store page links back to the full game, the demo in a player's library shows "a prominent link" to the full game's store page, and "You can direct customers to the full app from your demo by utilizing the Steam Overlay." Open the full game's store page in the overlay from your button, so the player can wishlist without leaving the demo.

Do not end on a hard cut to the main menu. That throws away the moment when interest is highest. A plain ask on the last screen, with one button, is enough.

Sources: Steamworks Demos documentation, read 19 November 2026; our practice.

How do you test the length before the fest?

Watch new players, then measure a larger group. Zukowski suggests live playtests with new players to find where they get confused or stuck, and an analytics event for each phase of the demo to find the drop. In his examples, one developer found a bug this way, and the developer of Tower Escape set the default difficulty to easy and saw the median rise from 15 to 42 minutes.

Steam Playtest is a good place for the larger group. Valve calls it "a free, low-risk way to get playtesting data for their game," and by default players request access and you grant it in batches. Run it before the demo goes public, so the demo launches at the right length.

  1. Write down the length you designed for, and the point where the demo ends.
  2. Watch three to five new players play without help. Note every place they pause, ask or quit.
  3. Add one event per phase of the demo, and one when the end screen shows.
  4. Run a Steam Playtest with a few hundred players. Read the median playtime and the share who reach the end screen.
  5. Cut or fix the phase with the biggest drop, ship a new build and compare the two builds.
  6. Stop when most testers reach the end screen. Then submit the demo at least 3 weeks before the fest.

Sources: How To Market A Game (26 October 2022, read 19 November 2026); Steamworks Steam Playtest documentation, read 19 November 2026.

Which demo length fits your game?

Pick the row that matches your demo, build to it, and let the playtest numbers move you. The table is our suggestion, built from the sources above; no study tests these exact lengths.

A starting demo length by kind of game. Our suggestion, not Valve guidance.
Kind of demoContent to buildWhere to endWhat to watch
Linear: story, puzzle, platformer, horror30 to 60 minutes of content for a first time playerAfter the first big reveal, boss or scareShare of players who reach the end screen
Replayable: roguelike, deckbuilder, autobattlerOne or two full runs, with content capped instead of timeAfter a run ends, with the locked content in viewRuns per player and median playtime
Loop: cozy, farming, managementOne full loop or in game day, about 20 to 40 minutesAt a moment of calm success, such as a first harvestShare of players who finish the first loop
Co-op or party gameOne short match or level that a group can finishAfter the first match, with the full mode list shownSessions per player, not minutes
Any demo with a median under 10 minutesShorter, not longer: cut the opening firstKeep the end where it isThe step where most players quit

Sources: our decision table, based on How To Market A Game (2022, 2024), Game Maker's Toolkit (2025), GameDiscoverCo (2025, 2026) and Steamworks documentation, all read on 19 November 2026.

How do you measure the result?

Manalens is the analytics and growth engine built for games. With the Godot or Unity SDK in the demo build, game_open, session_start and session_end (with its duration in seconds) are sent for you, so the Game KPIs board shows session length and sessions per daily active player for the demo. Set the build value to a demo name, so demo play stays apart from other builds. Add level or checkpoint events and a wishlist button click event to see a funnel from the first launch to the end screen.

A click on your button is not a wishlist: Steam reports wishlists, and it does not say which player made one. Upload the Steamworks UTM report to see wishlists from tagged links, and with a Steamworks Financial API key the automatic Steam sync (in early access) records all wishlist adds and deletes per day and country. Compare those days with the demo launch and each build.

Sources: Manalens SDK, app events, metrics and Steam documentation.

Frequently asked questions

Is there a minimum or maximum length for a Steam demo?

No. Valve says demos can range in size and scope, and its Next Fest page says there are no requirements on the length of the demo. Valve only asks you to give players enough to get excited without giving away everything the game has to offer, and to keep the quality high.

What is a good median playtime for a Steam demo?

In How To Market A Game's 2022 benchmark of 130 demos, the median was 14 minutes, with 22 minutes at the 70th percentile. Demos of games that later reached the top revenue tiers had medians of 38 and 65 minutes. Under 10 minutes put a demo in the bottom 30%.

Should a roguelike demo be longer than a story demo?

It usually plays longer. In the 2022 benchmark, replayable demos had a median of 18.5 minutes against 14 for linear ones. Cap a roguelike demo by content, such as a set of cards or one mode, not by time, and let players start more runs inside that limit.

Should the demo end with a wishlist button?

Yes. End on a high point and show an end screen with the release window and a Wishlist on Steam button. Valve says you can send players from the demo to the full game with the Steam Overlay, and Steam also links the demo in the library to the full game's store page.

Can I change the demo length after Next Fest starts?

Yes. Valve says you can keep updating the demo after it passes build review, but it will not review it again, so test each build yourself. The one wishlist notification for a demo can be sent only within 14 days of its first release, so settle the length before that.

Sources and implementation