SambaNova (Whisper)

​

Overview

SambaNovaSTTService provides speech-to-text capabilities using SambaNova’s hosted Whisper API. It offers high-accuracy transcription with minimal setup requirements. The service uses Voice Activity Detection (VAD) to process only speech segments, optimizing API usage and improving response time.

​

Installation

To use SambaNovaSTTService, install the required dependencies:

pip install "pipecat-ai[sambanova]"

You need to set up your SambaNova API key as an environment variable: SAMBANOVA_API_KEY.

​

Configuration

​

Constructor Parameters

​

Input

The service processes audio data with the following requirements:

• PCM audio format.
• 16-bit depth.
• Single channel (mono).

​

Output Frames

The service produces two types of frames during transcription:

​

TranscriptionFrame

Generated for final transcriptions, containing:

​

ErrorFrame

Generated when transcription errors occur, containing error details.

​

Methods

​

Set Model

await service.set_model("Whisper-Large-v3")

See the STT base class methods for additional functionality.

​

Language Support

SambaNova’s Whisper API supports a wide range of languages. The service automatically maps Language enum values to the appropriate Whisper language codes.

The service will automatically detect the language if none is specified, but specifying the language typically improves transcription accuracy.

For the most up-to-date list of supported languages, refer to the SambaNova’s docs.

​

Usage Example

from pipecat.services.sambanova.stt import SambaNovaSTTService
from pipecat.transcriptions.language import Language

# Configure service
stt = SambaNovaSTTService(
    model="Whisper-Large-v3",
    api_key="your-sambanova-api-key",
    language=Language.EN,
    prompt="Transcribe the following conversation",
    temperature=0.0
)

# Use in pipeline
pipeline = Pipeline([
    transport.input(),
    stt,
    llm,
    ...
])

​

Voice Activity Detection Integration

This service inherits from SegmentedSTTService, which uses Voice Activity Detection (VAD) to identify speech segments for processing. This approach:

• Processes only actual speech, not silence or background noise.
• Maintains a small audio buffer (default 1 second) to capture speech that occurs slightly before VAD detection.
• Receives UserStartedSpeakingFrame and UserStoppedSpeakingFrame from a VAD component in the pipeline.
• Only sends complete utterances to the API when speech has ended.

​

Metrics Support

The service collects the following metrics:

• Time to First Byte (TTFB).
• Processing duration.
• API response time.

​

Notes

• Requires valid SambaNova API key.
• Uses SambaNova’s hosted Whisper model.
• Requires VAD component in transport.
• Processes complete utterances, not continuous audio.
• Handles API rate limiting.
• Automatic error handling.
• Thread-safe processing.

​

Error Handling

The service handles common API errors including:

• Authentication errors.
• Rate limiting.
• Invalid audio format.
• Network connectivity issues.
• API timeouts.

Errors are propagated through ErrorFrames with descriptive messages.