This new AI model hears your tone, senses your mood, and talks back like a real human. Siri could never. (www.businessinsider.com)

Inworld AI has unveiled its new voice model, Realtime TTS-2, which enhances conversations with machines by interpreting vocal cues such as tone, pacing, and pitch to detect a user's emotional state in real time. This advancement allows the AI to adjust its voice and delivery dynamically, creating interactions that feel more natural and emotionally aware—a significant leap over existing models that primarily focus on accurate speech without contextual understanding. By resolving the emotional layer of communication, Inworld aims to make AI interactions as engaging as human dialogues, opening the door for broader usage across areas like customer service, healthcare, and education. Key to TTS-2's innovation is its ability to maintain contextual awareness throughout a conversation, drawing on historical dialogue to respond appropriately to various emotional cues. During a live demonstration, CEO Kylan Gibbs showcased how the model can shift its tone based on the conversational context, delivering nuanced responses that reflect emotional intelligence—an area largely overlooked by current voice AI technologies. Positioned as a foundational tool for developers through an API rather than a direct consumer product, Inworld is focused on empowering creators to build their own applications, thereby sidestepping competition with its clients and fostering a robust ecosystem for emotionally-aware AI interactions.