Ego Revolutionizes Gaming with Interactive Characters Powered by Cartesia

"Gaming has always been where communities form - from my generation's World of Warcraft and Runescape to today's Roblox and Minecraft. As games evolve into social platforms, AI characters need to feel genuinely human in both their responsiveness and emotional depth. Cartesia's technology, with its ultra-low latency, natural voices, and precise emotional control, helps us create truly immersive worlds where AI characters feel alive and authentic."

— Peggy Wang, Co-Founder and CTO, Ego

Ego is an AI-native simulation engine where users can create and share 3D animated characters, worlds, and game scripts using natural language. Ego’s vision is revolutionary for gaming - imagine being able to prompt the most popular games today like Minecraft or Animal Crossing into existence through simple commands.

Ego’s founders bring deep expertise in both gaming and AI. CEO Vishnu Hari, a former Top 500 Overwatch player, previously led product teams at Facebook AI Applied Research (FAIR) and the Meta Horizon Scripting team. CTO Peggy Wang, who shipped ML algorithms for Meta's AR Avatar face tracking, brings experience from Stanford AI Research and Lyft Level 5 in autonomous behavior planning algorithms for robotics.

Partnering with Cartesia has accelerated Ego's vision of allowing gamers to spawn the 3D world they imagine, complete with AI agents and Non-Player Characters (NPCs) that display human-like behavior. This partnership was recently showcased with the launch of Thrall, a game mod for the Viking-themed survival game Valheim. Using Cartesia's ultra-low latency voice technology (<90 ms) and advanced voice design capabilities, Thrall responds to players with human-like reactions, emotional awareness, and natural interactions.

Check out the launch tweet from Ego here.

The Role of Voice AI in Immersive Gaming

Until recently, open-world gaming faced three major limitations:

• NPCs were limited to rigid, pre-programmed behaviors without true agency
• Creating game interactions required extensive coding expertise
• Developing and scaling 3D assets and environments was prohibitively resource-intensive

The proliferation of generative AI has changed this landscape dramatically. Ego saw an opportunity to democratize immersive gaming by enabling anyone to create rich gaming experiences through natural language—similar to filling out a character sheet in Dungeons & Dragons.

Voice AI enhances two critical elements in modern gaming:

1. Non-Player Characters (NPCs): The computer-controlled characters that populate game worlds—from shopkeepers to quest-givers
2. AI Companions: Intelligent agents that can make decisions and interact like human players

Both require voices that can match their intelligence and adapt to emotion and context in real-time.

The Challenge

When evaluating voice providers, Ego sought to solve several critical requirements:

1. Achieving natural, real-time voice interactions between players and AI companions
2. Generating contextually appropriate emotional responses
3. Supporting multiple languages for international players
4. Creating distinct voices for different character personalities
5. Maintaining low latency for seamless gaming experiences

The Solution: Sonic

Gaming demands voices that can match dynamic player interactions. Ego chose Cartesia's Sonic model for its distinctive capabilities:

1. Emotion-First Design: Unlike traditional text-to-speech models, Sonic enables AI companions to express a full spectrum of emotions—from curiosity to combat urgency—making every interaction feel authentic. Alternate providers typically have some form of emotion recognition pre-trained in their models, but contextual awareness falls flat. Cartesia gives customers fine-grained control over emotions without having to regenerate the voice over and over again.
2. Gaming-Optimized Latency: Powered by Cartesia's breakthrough SSM architecture, Sonic delivers industry-leading 90ms latency—essential for maintaining immersion in real-time gaming environments.
3. Unlimited Dynamic Character Voices: With Cartesia's instant voice cloning capability, each AI character maintains its unique voice characteristics, trainable with just 10 seconds of audio. This enables Ego to create diverse, memorable characters that resonate with players while maintaining consistent personalities.
4. Global Gaming Community: Ego’s games reach a global audience across 14 other languages in addition to regional accents, ensuring authentic character interactions worldwide.
5. Voice Prompting: Cartesia's voice changer allows precise control over dialogue delivery, replacing traditional studio sessions with efficient voice-prompted commands. Developers can record a sample with the desired style, emotion, and prosody, then apply it to any character voice—enabling natural voice interactions without the need for repeated studio sessions.

Results

Ego's launch of Thrall represents a breakthrough in NPC design—players can use natural voice commands to direct it in performing any task a human player would do, from gathering resources to engaging in combat. The AI companion responds with contextually appropriate emotions, celebrating successful hunts or acknowledging commands with human-like reactions. Most importantly, players can customize Thrall's personality and behavioral traits, creating a truly personalized gaming companion.

The partnership represents a significant step toward Ego's vision of democratizing game creation and enabling truly interactive gaming experiences where AI characters feel alive and responsive.

"As an avid gamer, I've been amazed by Ego's breakthrough in creating truly responsive gaming experiences. Previously, NPCs and game characters were limited by rigid, pre-programmed behaviors. Sonic brings these characters to life with natural personalities and expressiveness. I'm excited to contribute to Ego's vision for the future of interactive gaming."

— Karan Goel, CEO, Cartesia