Planned Release Date: September 2026

About This Game

Paleo Dive

Dive into a world long lost beneath the waves. Paleo Dive is a calm VR diving adventure that takes you through prehistoric underwater landscapes — from vibrant coral reefs and dense kelp forests down into the quiet darkness of the deep sea.

Explore. Collect. Discover.

Glide through the water using your own arm movements, keep an eye on your oxygen supply, and make it back to base in time. Along the way, you'll uncover finds that unlock new diving areas and let you gradually upgrade your equipment.

Features

• Intuitive swimming driven by natural arm motion

• Several handcrafted dive sites, each with its own atmosphere

• Oxygen and return-to-base management for tension on every dive

• Collection and upgrade system accessed through a wrist-mounted smartwatch

• Multiple save slots so you can explore at your own pace

VR Only – OpenXR

Paleo Dive is playable exclusively in virtual reality and supports headsets through the open OpenXR standard.

AI Generated Content Disclosure

The developers describe how their game uses AI Generated Content like this:

In accordance with digital platform transparency guidelines, it is hereby disclosed that advanced generative computational models and machine-learning frameworks have been selectively integrated into the multifaceted production pipeline of this interactive software application. Specifically, algorithmic synthesis was deployed during the asset procurement phase to facilitate the generation of various graphic user interface (GUI) elements, incorporating specific iconography, non-diegetic symbols, and visual indicators. Furthermore, the auditory landscape features synthesized sound effects derived through prompt-based neural network modulations.

Regarding the structural architecture, the underlying source code was developed utilizing automated algorithmic assistance tools for debugging, syntax optimization, and routine logical construction, while the three-dimensional environment comprises assets where generative workflows intermittently assisted in the final topology refinement, polygonal optimization, and the algorithmic projection of procedural and raster-based texturing layers for a non-exhaustive subset of 3D assets. Lastly, contextual textual localization, narrative phrasing, superficial UI layout configurations, and various minor supplementary asset adjustments were subject to iterative large language model oversight and machine-assisted content generation.