Melody's Escape 2
Melody's Escape 2 is making a strange bet on the music you already own
The original Melody's Escape lived in a quiet corner of Steam for years. One developer. A little blue runner. You pointed it at your music folder and trusted the algorithm to figure out the rest. No DLC packs. No subscription. No label tie-ins. The sequel, Melody's Escape 2, hits Steam in June 2026, and Icetesy is doubling down on the exact thing that made the first one a cult pick. Your library. Your problem. Your flow state.
In 2026, that's a weird hill to stand on.
The rest of the genre spent the last decade walking the opposite direction. Curated tracks, licensing deals, season passes. Rhythm games are basically a music-licensing business with a game bolted on. Icetesy is selling the inverse pitch: bring your own FLACs, OGGs, M4As, and the engine will build the level on the fly. The Steam page even warns you, in plain English, that streaming services won't work because they can't. You almost never see that kind of honesty in a store listing.
The algorithm is the game
Icetesy keeps hammering one phrase in the marketing copy. "Human-made" analysis algorithms. That's a pointed line in a year when every other audio tool on the planet is gluing a neural network to the side of the box and calling it innovation. Tempo detection, intensity curves, vocal isolation, all hand-coded.
Whether that's a flex or a limitation depends entirely on what you feed it. The first game was great with rock, electronic, anything with a clean kick drum. It got considerably more confused by jazz, ambient, drone, anything with a shifting time signature. The sequel claims "vastly improved" detection across the board. Only release-week stress testing from the community will actually settle that one.
The difficulty tiering, though, is genuinely smart. Relaxing mode is one button. Medium uses four colored inputs. Intense pushes you to eight, which is where the game stops being a chill flow exercise and starts being a real reaction test. There's also a custom difficulty mode, which tells you Icetesy knows its audience. The hardcore players will absolutely build their own profiles and trade them around.
What people are actually doing with it
Look at what creators are already posting from the beta and the pattern shows up fast. They aren't playthroughs in the normal sense. They're perfect runs. Specific songs. "Free Bird Solo." "Year Zero." System of a Down's "Soldier Side."
The interesting bit is that the videos aren't really about the game. They're about the pairing of song and player, with the game as the conduit. That's a totally different content loop than a standard rhythm title, where everybody is grinding the same DLC pack the same week.
Each upload is, in effect, a music video the player made by being good at the game. That engine runs forever as long as people keep finding tracks they want to beat themselves up against. The Steam Workshop cosmetic skins are a side dish. The real user-generated content here is the music itself.
14 hand-mapped tracks, and why that number is the point
The base game ships with 14 hand-mapped levels across various styles. By rhythm-genre standards that's a tiny number, and it's pretty clearly meant as a demo reel. Proof of what the algorithm can do when a human carefully tunes the chart.
The pitch isn't "play these 14 songs forever." It's "look, the engine works, now go raid your hard drive." If you've got a serious local collection, a few thousand tracks of personal history sitting on a NAS somewhere, that proposition is honestly compelling. If you've lived entirely inside Spotify since 2015, it's a non-starter, and Icetesy says so right on the store page.
That self-selecting audience is basically the whole story of whether this thing survives commercially.
The marketing problem
A mass-market push for Melody's Escape 2 would set money on fire. The overlap between "people who still maintain a local DRM-free music library" and "people who play rhythm games" is small, passionate, and almost completely unreachable through generic Steam ads or influencer-network blasts.
You can't sell this to the casual mobile rhythm crowd. They don't have FLAC files. You can't sell it to a Beat Saber kid who only knows curated packs.
The realistic path for a studio this size is finding the people who already care about the specific intersection. Home-theater audiophile YouTubers who obsess over codecs. HTPC and retro-PC creators with massive ripped libraries. Synesthesia and music-visualization channels. Speedrun and perfect-run communities like the ones already posting clips. The rhythm purist tribe that grew up on osu! and Audiosurf. Reaching those creators one at a time, with builds in hand, is how Icetesy actually punches through launch week. Tooling like CreatorFetch is what makes that kind of hyper-targeted outreach tractable instead of a months-long DM scramble.
A verdict, sort of
Judging a rhythm game that hasn't shipped is hard, especially one whose quality depends on songs the reviewer hasn't picked yet. So take this for what it is.
Icetesy is making the same bet again, with sharper tools and a clearer read on who actually wants the thing. The first Melody's Escape survived more than a decade on word of mouth from people telling their friends, "no, really, try it with your own music." If the audio analysis holds up on the weird stuff, prog metal in 7/8, sparse acoustic recordings, anything with tempo shifts, the sequel will earn the same slow-burn loyalty.
And if it doesn't? It'll still be the best version of itself the genre has, because almost nobody else is even trying to build this specific thing anymore. That's either a moat or a warning sign, depending on how you read the market.