Inspiration
We wanted to explore how voice input could become the primary control mechanic in a high‑intensity game. Inspired by the “Split Second” theme, we aimed to merge speech recognition with instant reaction gameplay, creating a unique challenge where hesitation means failure. In Chant, every moment counts — enemies, obstacles, or events appear suddenly, and the only way to respond is by speaking the correct word or phrase instantly. The “Split Second” theme is woven into every interaction: the faster and more accurately you speak, the better your chances of survival. A single delayed or incorrect chant can change the outcome, making timing and precision the heart of the experience.
Gameplay
Players face a series of rapid‑fire challenges on screen. Instead of using a keyboard or mouse, they must speak specific commands into their microphone to perform actions — jump, attack, defend, or interact. The game listens in real time, rewarding quick, correct responses and penalizing hesitation or mistakes. Difficulty ramps up as reaction windows shrink, pushing players to think and speak faster under pressure.
How It Was Built
We developed Chant in Unity and exported it as a WebGL build for browser play. The voice recognition system uses the Web Speech API integrated into Unity via JavaScript plugins, enabling low‑latency speech detection. Visuals and audio were optimized for quick loading, and the game was deployed using Tencent Cloud EdgeOne for global reach and minimal input lag.
Built with
1. EdgeOne Pages for hosting and deploying the WebGL build. 2. Global Content Delivery Network (CDN) to ensure low‑latency gameplay worldwide. 3. Security Acceleration to protect the game from malicious traffic and ensure stable performance. 4. Edge Functions for lightweight serverless processing of voice command validation and leaderboard updates.

