STT Providers

​

What are STT Providers?

• Real-time transcription - Convert speech as it happens
• Multiple languages - Support for various languages
• High accuracy - State-of-the-art transcription models
• Streaming support - Low-latency processing

​

Available Providers

​

Choosing a Provider

​

Deepgram

• High accuracy requirements
• Real-time streaming
• Multiple language support
• Fast response times

• DEEPGRAM_API_KEY environment variable

​

Google Cloud STT

• Google Cloud ecosystem integration
• High accuracy
• Enterprise deployments
• Vertex AI integration

• GOOGLE_STT_API_KEY OR GOOGLE_STT_CREDENTIALS environment variable

​

Configuration

​

Basic Configuration

from kuralit.server.agent_session import AgentSession

# Using Deepgram
agent = AgentSession(
    stt="deepgram/nova-2:en-US",
    # ...
)

# Using Google Cloud STT
agent = AgentSession(
    stt="google/en-US",
    # ...
)

​

Next Steps

• Deepgram → - Deepgram provider documentation
• Google Cloud STT → - Google Cloud STT documentation
• Voice → - Learn about voice streaming
• STT → - Understand Speech-to-Text