CODE EXIT’s New Playtest Is Live, and It Rebuilds the Demo’s Biggest Problems
CODE EXIT’s new Steam playtest is live from September 28 to October 19, and this one looks worth another go. Phase 8 hasn’t simply reopened the previous demo and changed the wallpaper. The mission structure, level layouts, co-op objectives, death system and the way S.O.M.A. hunts players have all been reworked.
That gives the playtest an actual reason to exist.
What we cannot say yet is whether CODE EXIT is now better. The developers have changed several things that needed changing. Whether they actually fixed them is the interesting part.
If CODE EXIT has awakened your unhealthy interest in co-op expeditions where everything goes wrong simultaneously, Murky Divers has already taken that experiment underwater. See whether the screaming is worth it.
CODE EXIT has basically put an AI in charge of an entire city, which went exactly as expected
CODE EXIT is a 1-4 player co-op survival horror game set underneath Pantheon, a city controlled by an integrated AI called Minerva.
Minerva has, unsurprisingly, not used this responsibility to improve refuse collection.
You descend into hostile facilities, complete objectives, recover C.O.D.E. and try to get back out while avoiding S.O.M.A., the mechanical thing wandering around making the whole expedition considerably less like archaeology and considerably more like being trapped in a warehouse with industrial equipment that has developed opinions.
That basic setup hasn’t changed.
A lot around it has.
The old demo split progression across three sessions. The new playtest rolls that into one run.
Objectives have been simplified, level layouts now change between sessions, there are new areas requiring multiple players, the death system has been rebuilt around MiRO, Mental Contamination has been added, and players now have a Scanner and MiRO shop.
Individually, some of those sound like patch-note furniture.
Together, they start changing what a run actually is.
Especially because S.O.M.A. is no longer simply there from the beginning doing robot murder laps around the premises.
S.O.M.A. now sleeps until somebody inevitably kicks a bin
S.O.M.A. starts dormant.
Movement and noise can wake it.
This is a magnificent idea for a four-player co-op horror game because four-player co-op contains roughly the same amount of operational discipline as a stag party entering an IKEA.
Three people will crouch through a corridor.
Nobody moves.
Nobody speaks.
The fourth player runs into a chair.
Something metallic crashes onto the floor.
“Who the fuck did that?”
And somewhere in the darkness, several tonnes of machinery open their eyes because Kevin has once again failed the basic human challenge of walking.
There is an important mechanical point buried inside Kevin’s disciplinary hearing.
Phase 8 says it has also improved feedback around S.O.M.A.’s detection. That matters enormously. A horror pursuer needs rules, even if the player never gets handed those rules on a laminated card.
If you understand absolutely nothing about how it finds you, eventually the monster stops feeling intelligent. It just feels like the game secretly owns a button marked release robot directly behind player.
But explain everything too clearly and the monster becomes equipment.
You stop fearing it and start calculating it.
“Fourteen metres, line of sight broken, carry on.”
Congratulations. The horrifying machine from the depths of Pantheon is now a parking restriction.
CODE EXIT needs to sit somewhere between those two.
Tired of being chased by the homicidal machinery? SAW: Genesis lets you change departments and start running the murder dungeon yourself. HR may have questions.
Override might be the clever bit
The new Override system is probably the change I’m most curious about.
Minerva can periodically alter the rules governing S.O.M.A.’s pursuit.
That could be properly nasty.
A large part of surviving any horror enemy is learning it. You die, you understand something, you try again. Eventually the unknowable horror becomes Barry from Security, and you know Barry checks Corridor B every 40 seconds and apparently cannot see around filing cabinets.
Override threatens that process.
If the pursuit conditions can change during a run, yesterday’s clever little survival rule can suddenly become today’s reason your body is being peeled off a wall.
Add randomized layouts and CODE EXIT could make knowledge useful without allowing knowledge to completely sterilise the horror.
Could.
That word still needs to remain bolted firmly to the floor.
Phase 8 also describes Minerva as adapting to player behaviour, but we do not yet know how convincing that feels in practice. Games have been promising intelligent enemies for decades. Usually this lasts until the allegedly revolutionary AI spends four minutes trying to negotiate a doorway.
So that claim can wait.
The interesting thing we can actually verify is that the rules around the pursuer now have more room to change.
The developers have at least diagnosed the right patient
What makes this playtest more interesting than another “come try our game again” announcement is that Phase 8 has talked about specific problems with Demo 1.0.
The previous puzzle structure could become too dense.
Co-operation could feel artificially forced.
S.O.M.A.’s detection could be unclear.
Repeated sessions could become repetitive.
Those aren’t tiny balance complaints. They go straight through the middle of what CODE EXIT is supposed to be.
And the changes now being tested line up rather suspiciously well with them.
Objectives are simpler.
Layouts are randomized.
Co-op spaces have been redesigned.
S.O.M.A.’s detection feedback has been changed.
Override adds uncertainty.
That still doesn’t prove anything has been solved, but at least we are not watching a developer treat a broken leg by adjusting the hat.
There is a visible relationship between the problems and the work.
I like that.
Especially because replayable co-op horror can die very quickly once the horror has been solved. The first time you enter a corridor and something starts scraping metal in the darkness, your brain immediately begins writing a will.
The fifteenth time, somebody says, “He always comes from the left.”
And suddenly your unknowable mechanical nightmare is a bus timetable.
CODE EXIT seems very aware that this is the enemy.
Not S.O.M.A.
Familiarity.
So, should you play the new CODE EXIT playtest?
If co-op survival horror is your thing, yes.
The previous demo had positive reception, and LINE Games says it reached the top spot on Epic Games Store’s Top Demos chart. I’d still treat that ranking as the publisher’s claim rather than something handed down from the mountain on stone tablets, but there was clearly enough interest around the original demo to make this second attempt worth watching.
More importantly, this isn’t simply the same demo coming back for another lap.
Phase 8 has rebuilt enough of the experience that there is now something useful to test.
Does random generation actually keep repeated runs fresh?
Does Override make S.O.M.A. less predictable without making the game feel arbitrary?
Does Mental Contamination matter?
Does the improved detection feedback help you understand what went wrong without turning the monster into a household appliance?
And does co-op make Pantheon more tense, or does four-player survival horror once again become an elaborate experiment to discover which of your friends should never be trusted near a door?
The playtest runs until October 19.
The developers have already shown they know where several of the old problems were.
Now CODE EXIT has to prove it knows what to do with them.
For horror at the opposite end of the corridor, The Crimson Maid trades angry robots for a gorgeous gothic mansion and then quietly forgets to put enough game beneath the floorboards. Enter at your own risk.
CODE EXIT FAQ
When is the CODE EXIT playtest available?
The current Steam playtest runs from September 28 to October 19, 2026.
Is CODE EXIT free to play?
The playtest is free to join on Steam. The full game is planned for 2027.
Can you play CODE EXIT solo?
Yes. CODE EXIT supports 1-4 players, although some areas in the new playtest are specifically designed around co-op interaction.
What changed in the new CODE EXIT playtest?
Phase 8 has reworked the session structure, simplified objectives, added randomized layouts, introduced Mental Contamination and MiRO systems, changed S.O.M.A.’s behaviour and added the new Override mechanic.
Is the new CODE EXIT playtest better than the old demo?
Too early to say. The developers have clearly targeted several problems from Demo 1.0, but whether those changes actually improve the game needs proper hands-on testing.
Comments