Navana STT

The Navana STT provider enables your agent to use Navana's Bodhi speech-to-text models, which are highly optimized for a variety of Indian languages and accents.

Installation

Install the Navana-enabled VideoSDK Agents package:

pip install "videosdk-plugins-navana"

Authentication

The Navana plugin requires a Customer ID and an API Key from your Navana Bodhi account.

Set both NAVANA_API_KEY and NAVANA_CUSTOMER_ID in your .env file.

Importing

from videosdk.plugins.navana import NavanaSTT

Example Usage

from videosdk.plugins.navana import NavanaSTT
from videosdk.agents import CascadingPipeline

# Initialize the Navana STT model
stt = NavanaSTT(
    api_key="your-navana-api-key",
    customer_id="your-navana-customer-id",
    model="en-in-general-v2-8khz",
    language="en-IN"
)

#  Add stt to cascading pipeline
pipeline = CascadingPipeline(stt=stt)

note

When using a .env file for credentials, don't pass them as arguments to model instances. The SDK automatically reads environment variables, so omit api_key, customer_id, and other credential parameters from your code.

Configuration Options

• api_key: Your Navana API key (required, can also be set via NAVANA_API_KEY environment variable).
• customer_id: Your Navana Customer ID (required, can also be set via NAVANA_CUSTOMER_ID environment variable).
• model: The Navana STT model to use (e.g., "en-in-general-v2-8khz", "hi-general-v2-8khz").
• language: The language code for transcription (e.g., "en-IN", "hi-IN").

Additional Resources

The following resources provide more information about using Navana with the VideoSDK Agents SDK.

• Python package: The videosdk-plugins-navana package on PyPI.

• GitHub repo: View the source or contribute to the VideoSDK Navana STT plugin.

• Navana Docs: Navana's official streaming API documentation.