Square Enix will run a second closed beta test (CBT 2) for KILLER INN on Steam from October 3 to 13, 2025, giving players another window to test the 24-player murder-mystery action title. The test arrives three months after a July beta that focused primarily on server resilience, and it aims to validate sweeping balance changes, new systems, and matchmaking improvements. Players who joined the first test are automatically opted in, while new testers can sign up through the Steam playtest program or earn on-site codes at the Tokyo Game Show.

KILLER INN blends third-person action combat with social deduction at an unprecedented scale. Up to 24 players are split into a minority team of hidden killers (Wolves) and a majority of survivors (Lambs). Instead of relying on vote-kicking, suspects are confronted through real-time combat, while survivors collect forensic evidence—hair, fingerprints, clothing fibers—to identify murderers. The hybrid design promises cinematic tension, but early feedback revealed deep balance and moderation challenges. CBT 2 is the proving ground for whether the developer, Tactic Studios, can tune the asymmetry without breaking the investigative core.

Beta Access and Key Dates

CBT 2 runs from October 3 through October 13, 2025, exclusively on Steam. Square Enix has published exact start and end times for each region in its announcement. Prior participants from the July closed beta will retain access automatically—they need only ensure the Steam client is updated. Newcomers can request access via the Steam Playtest button on the store page; invitations will be sent in waves before the test begins. Tokyo Game Show attendees who demo the game at Square Enix’s booth can receive physical invitation codes, adding an in-person recruitment channel.

The extended two-week window is longer than the July test, designed to stress servers over sustained periods and capture a wider sample of cross-region matchmaking. Testers should plan sessions across different time zones to surface latency and queue-fairness issues that shorter tests often miss.

System Requirements: A Demanding Client

Square Enix published minimum and recommended PC specs for the beta client, and they’re notably steep. Both tiers require Windows 10 64-bit or Windows 11, 16 GB of RAM, and an SSD with 75 GB of free space. The minimum GPU target is an AMD Radeon RX 5500 XT (8 GB) or NVIDIA GeForce GTX 1060 (6 GB), aiming for 1920×1080 at 30 FPS. The recommended spec bumps the CPU to a Ryzen 5 5500 or Intel Core i7-9700K / i5-10600, and the GPU to an RX 5700 or RTX 2060 SUPER, targeting 60 FPS at the same resolution.

Two constraints stand out: the fixed 16 GB RAM floor suggests heavy memory usage from 3D spatial audio, high-fidelity assets, and background recording features, so streamers may want 32 GB. The 75 GB install footprint and mandatory SSD point to extensive asset streaming, likely of detailed character models, large maps, and spatial audio data. Users with mechanical hard drives will face severe stuttering and load times, potentially disqualifying them from effective play. Testers should free at least 100 GB on their SSDs to account for shader caches and write overhead.

What’s New: Balance Pass and System Additions

The July beta was a foundational network test. CBT 2 is explicitly an iterative design test. Square Enix’s developer messaging highlights three focus areas: balance tuning, new system integration, and structured feedback collection. While the studio hasn’t shared full patch notes, the forum community has catalogued the most pressing pain points from the first test, and they give direction on what changes might land.

Pacing and the 24-Player Problem

Large-scale social deduction easily tips into chaos. In the July beta, matches often devolved into wide-scale brawls within minutes, robbing the investigative layer of oxygen. Killers could snowball early if they secured weapons quickly, while survivors lacked non-lethal counterplay. CBT 2 must demonstrate that loot distribution and combat lethality have been reined in. Expect weapon spawn timers, equipment cooldowns, and perhaps an early-game grace period that preserves deduction time. The addition of more “neutral” objectives—quests and lootable chests—may also spread players across the map, reducing dense firefights.

Forensic Evidence Tuning

The forensic system is KILLER INN’s key differentiator. In theory, hair, prints, and fibers at murder scenes let Lambs build a logical case instead of relying solely on persuasion. But in practice, the July test saw extremely noisy clues: false positives that condemned innocent players, or overly silent scenes that gave no leads. CBT 2 likely includes a signal-to-noise pass. Clues should be informative without being deterministic, preserving bluffing as a valid Wolf strategy while giving observant Lambs a real advantage.

Combat vs. Deduction: A Fork in the Road

When a culprit is cornered, the game shifts into direct combat rather than a vote. This raises the skill ceiling but risks alienating players who excel at deduction but not at aiming. The July test favored combat-heavy players—even a Wolf caught red-handed often won duels. For CBT 2, testers should look for environmental counters (traps, barricades), specialized non-combat abilities (tracking, forensic analysis), and weapon lethality adjustments. If Lambs can neutralize a stronger foe through tactics rather than twitch skill, the hybrid promise holds. Otherwise, the game collapses into a battle royale with extra steps.

3D Spatial Voice: Immersion’s Double-Edged Sword

KILLER INN features proximity-based 3D voice chat. Players hear others realistically, fading with distance and direction. It’s a powerful immersion tool—you can overhear whispered conspiracies or a Wolf’s panicked lie. Yet it’s also a moderation nightmare. Open-mic proximity chat invites harassment, hate speech, and griefing at a scale most social-deduction games avoid by limiting voice to meetings or parties.

The July beta had minimal safeguards. CBT 2 must show that Square Enix and Tactic Studios are serious about community health. At minimum, public servers should default-mute voice with a clear opt-in, provide per-player volume and mute controls during matches, and integrate a robust reporting system that captures voice clips or timestamps. Behavioral matchmaking—isolating repeat offenders into low-priority queues—would be a strong sign. Without these, the proximity voice feature will drive away the very players it’s meant to enhance.

Matchmaking, Servers, and Network Stress

CBT 2 is longer specifically to test endurance and cross-region connectivity. The July test revealed sporadic disconnects and lopsided latency in mixed-region matches. This time, Square Enix will monitor reconnection logic, host migration stability, and matchmaking fairness. Dedicated servers are likely, but the protocol hasn’t been detailed. Testers should note packet loss, observed tick rates, and whether disconnects are clean (rejoinable) or fatal (match-ending).

For PC players, network best practices apply: wired Ethernet, updated drivers, and closing background bandwidth hogs. The game’s 75 GB SSD requirement suggests aggressive asset streaming during gameplay, which can saturate I/O and cause lag if the drive isn’t fast. NVMe SSDs are recommended.

The Looming Monetization Question

Square Enix has not announced a business model for KILLER INN. The CBT materials focus purely on technical and design goals. That silence is standard for early betas, but the community is already speculating about battle passes, cosmetic shops, and potential pay-to-win items. Pay-to-win would be catastrophic in a deduction game—any advantage that undermines the information economy (e.g., paid consumables that hide forensic evidence) would fracture fairness. The more likely path is cosmetic monetization: character skins, weapon skins, emotes. CBT 2 probably won’t include any store, but testers should watch for progression systems that could later be monetized, such as experience boosters or equipment unlocks that bleed into match power.

Practical Advice for CBT 2 Testers

To get the most out of the beta and provide actionable feedback, players should prepare their systems and approach sessions intentionally.

  • Free up at least 100 GB on an NVMe SSD. Install the client well before the test starts.
  • Update GPU drivers (NVIDIA/AMD) and confirm DirectX 12 functionality.
  • Use a wired internet connection to minimize packet loss.
  • Set up and test a microphone—adjust Windows and Steam voice settings. Join with friends to stress squad mechanics.
  • Play both roles (Lamb and Wolf) extensively to give balanced feedback on evidence systems and combat.
  • Keep a notepad handy to log timestamps of bugs, disconnects, and balance anomalies. Reproducible steps are gold for developers.
  • Monitor performance with tools like MSI Afterburner or Windows Game Bar. Note FPS drops, stuttering during asset loads, and any GPU/CPU utilization oddities.

High-quality bug reports and feedback directly influence post-beta patches. The more precise the data, the better the final game.

Community and Moderation: The Make-or-Break Factor

Social-deduction games live and die by community health. KILLER INN amplifies risk with its 24-player persistent voice chat. Toxicity can ruin matches instantly—one screaming troll or a coordinated griefing squad poisons the well. The developers need to implement a three-layered defense:

  1. Immediate controls: Mute, deafen, and report buttons accessible via a single hotkey or radial menu. Time-limited votekicks for obvious disruptors.
  2. Post-match enforcement: Detailed reporting categories with audio snippet attachments, plus human-reviewed temporary bans.
  3. Behavioral matchmaking: A hidden trust score that pairs habitual offenders together, quarantining them from the general pool.

CBT 2 is the first public stress test of these systems under load. The demographic will be smaller than an open beta, but bad actors still appear. How quickly and effectively Square Enix responds to early abuse reports will signal the studio’s commitment.

Strengths and Risks: A Pre-Beta Verdict

KILLER INN is conceptually bold. It grafts the tactile thrill of action combat onto the psychological depth of social deduction. At its best, it could produce stories players recount for years: the Wolf who fooled everyone until a lone Lambs reconstructed their trail of clues, then took them down in a tense skirmish. Square Enix’s backing and Tactic Studios’ experience (the team includes veterans with multiplayer pedigree) give it production muscle.

But the risks are proportional. Balancing 24-player asymmetry is a nightmare; most games avoid it entirely. The forensic system must work flawlessly or it becomes a random distraction. And the voice chat, while immersive, is a loaded gun. If the beta fails to demonstrate significant balance progress and robust moderation, the wider gaming press and content creators will brand it a toxic mess—and social-deduction audiences are unforgiving.

What to Watch During and After CBT 2

As CBT 2 unfolds, these indicators will reveal whether KILLER INN is on track:

  • Stability: Session stability, reconnection success rates, and matchmaking latency variance. If matches regularly crash or leave players stranded, launch confidence plummets.
  • Balance outcomes: The frequency of snowball victories—where one side wins overwhelmingly early—should decrease. Individual weapon dominance (e.g., a single sword outperforming all other tools) needs nerfing.
  • Evidence reliability: Forensic clues should lead to the correct killer more often than not without making bluffing impossible. The sweet spot is an evidence system that rewards diligence, not guesswork.
  • Moderation responsiveness: Count the minutes between a report and visible action. Even a generic “action taken” notification builds trust.
  • Community sentiment: Watch the Steam forums and Reddit. If top posts remain dominated by bug reports rather than toxicity complaints, the moderation layer is working.

Square Enix has not announced a launch window or final business model. Any claims about release dates or monetization are speculation until official confirmation. CBT 2 is a design audition—a checkpoint to see if the ambitious hybrid can be hammered into a stable, fair, and fun experience. For Windows PC players, it’s a rare chance to peer inside a high-stakes multiplayer experiment and help shape its future.