Publisher CMD Studios and developer Winter Crew have announced that side-scrolling Metroidvania Fallen Tear: The Ascension will launch in Early Access for PC via Steam on March 17. A full release is planned at a later date across PlayStation 5, Xbox One & Series X|S, Switch, Switch 2, and PC.

According to CMD Studios founder Stephen Manalastas, the Early Access launch marks an important milestone in the project’s evolution. The team intends to collaborate closely with its community—including longtime fans and Kickstarter backers—throughout the Early Access period to refine and expand the experience.
The Early Access version will offer between 15 and 20 hours of content, including 10 interconnected regions to explore, six main bosses, three optional bosses, and the opening act of the story. Additional updates are planned over the coming months as development continues toward the full release.
Players take on the role of Hira, a mysterious child tasked with confronting ancient gods in the world of Raoah. Exploration spans lush forests, sprawling cave systems, and shifting environments that evolve in response to Hira’s actions. A key gameplay system revolves around forming “Fated Bonds” with allies encountered throughout the journey. These bonds grant new abilities that enhance both combat and traversal, allowing players to unlock previously inaccessible paths and approach encounters with expanded strategic options.
The game integrates traditional Metroidvania structure with JRPG-inspired elements. Players can recruit party members across Raoah and unleash combined abilities through powerful Fated Rush attacks. Character progression includes skill upgrades and mastery systems that deepen combat mechanics, while enemy encounters emphasize pattern recognition and movement precision.
Visually, Fallen Tear: The Ascension features hand-drawn 2D animation and full voice acting, reinforcing its narrative-driven presentation. As players uncover hidden pathways and secrets embedded throughout the interconnected map, the world responds dynamically, encouraging revisitation and thorough exploration.
