The Inworld AI MCP server is an implementation of the Model Context Protocol that connects your AI agent and assistants like Claude, Cursor, etc directly to your Inworld AI account. It provides structured and secure access to speech, voices, models, and embeddings, so your agent can synthesize speech, transcribe audio, explore voices, compare models, and create text embeddings on your behalf.
- Speech synthesis: Have your agent turn short text into spoken audio using an available Inworld voice.
- Audio transcription: Direct your agent to transcribe short audio files and return the final text with timestamps and usage details.
- Voice discovery and previews: Let the agent browse available voices by language or other preferences and generate MP3 previews before you choose one.
- Model comparison: Instruct your agent to list supported models and compare their pricing, modalities, context sizes, and output limits.
- Text embeddings: Have your agent generate a complete embedding vector for each text input using an Inworld-hosted model.