Inworld

Inworld AI builds realtime voice models and LLM inference for consumer applications, including companions, learning, health and wellness, social apps, personal agents and entertainment. The platform hosts models such as Realtime TTS-2, Realtime TTS-2 Flash, and Realtime STT-1, and serves open models on a fleet the company operates across multiple regions.

The company describes its offering as designed for consumer AI, with fast first audio and low latency in speech workflows. It emphasizes an integrated approach where voice data and model inference work together through a single API surface, enabling developers to build interactive, voice-enabled experiences.

Inworld positions its products as suitable for teams building consumer-facing experiences, including games, education, wellness apps, and social companions. The documentation and playgrounds are presented as entry points for developers to explore voices, models, and inference capabilities.

The pricing and plans section indicates a tiered structure with On-Demand, Creator, Builder, Developer, Growth, and Enterprise options, each with different credits, limits, and features like custom voices, voice cloning, and HIPAA/BAA support.

Use cases

• Companions and social apps with real-time voice
• Learning, education and language practice
• Health and wellness coaching
• Games and interactive media with live dialogue
• Agentic workforce and enterprise agents