What Deep Worlds learned from its summer 2026 playtests
PvP called fun, onboarding called bad, and a plan to grow from 250 players to thousands: Deep Worlds' own July 2026 playtest report.
In early July 2026, Deep Worlds ran a playtest with real playtesters, and its CEO Aherys posted an unusually frank report while the session was still running. It is the most detailed picture of the game's state that the studio has published this year.
The report, point by point
Aherys sorted the early impressions into what worked, what needs work, and what is bad. Here they are as he wrote them, grouped the same way.
| Verdict | Feedback |
|---|---|
| Good | PvP is confirmed as very fun |
| Good | Performance is good for most players, lower configurations still need work |
| Good | Server performance is perfect for now |
| Good | Stability is good overall, with some crashes on the latest NVIDIA driver being worked on |
| Good | The Wendigo is scary for operators and very fun for the players controlling it |
| Good | Progression seems fun and rewarding |
| Good | Matchmaking issues are solved for most players |
| Needs work | Early progression for new players is very hard and not fun |
| Needs work | The Wendigo may be a bit too hard for early players |
| Needs work | Playing solo is less fun than in a team |
| Needs work | The game needs polish, and there is time for it |
| Bad | Onboarding is bad, and a tutorial is needed |
| Bad | Getting around the map is not very interesting yet, with a plan already in place |
| Bad | Operators lack tools against anomaly players, also planned |
Two of those "bad" points line up with what Deep Worlds said two months later. On 8 September, Aherys listed a game tutorial among the team's current priorities. And the reworked Ground Zero, with its new points of interest, is the studio's answer to a raid that felt too linear.
The last core feature
The report ended with the next milestone: anomaly and human progression, with skill abilities. Aherys called it the final core feature, after which he said development becomes a straight line of polishing, balancing and adding content before release. Deep Worlds showed a first look at that skill tree in August and September.
From 250 players to thousands
On 8 July, after another session, Aherys explained how playtests will grow. More than 200,000 people had asked to play, far too many to invite at once for a studio of Deep Worlds' size, and load balancing needs testing first. So the plan is to scale in steps:
- 250 players first
- then 500, then 1,000
- then 2,000, 5,000 and more
Discord members get priority, before playtests open on Steam as well, so the studio's advice is to join both. There is no date for the bigger waves.
On 2 August, Aherys added that the latest session had been a success for the servers. There were still some crashes, but he expected a new wave to follow quickly once stability on the matchmaking coordinator was sorted.
What it means if you want in
The route has not changed: the official Discord first, then Steam. The difference in 2026 is scale. The waves are meant to grow, and Discord members are first in line. Before you apply, check the system requirements, and see the release date page for where Early Access stands.
Sources
- Deep Worlds' official Discord (discord.gg/project-bl), #leaks-wip: posts by Aherys, CEO, on 2 July, 8 July, 2 August and 8 September 2026, and by Matavatar, Head Staff, on 15 May 2026 (helicopter extraction clip)
BL Guide, unofficial community coverage of Beautiful Light: how we check what we publish.