Whisper CPP for FastRTC

A PyPI package that wraps Whisper.cpp for speech-to-text (STT) transcription, compatible with the FastRTC STTModel protocol. This package provides efficient, CPU-based speech recognition using the optimized Whisper.cpp implementation.

Installation

pip install fastrtc-whisper-cpp

For audio file loading capabilities, install with the audio extras:

pip install "fastrtc-whisper-cpp[audio]"

For development:

pip install "fastrtc-whisper-cpp[dev]"

Usage

Basic Usage

from fastrtc_whisper_cpp import get_stt_model
import numpy as np

# Create the model (downloads from HF if not cached)
model = get_stt_model()

# Example: Create a sample audio array (actual audio would come from a file or mic)
sample_rate = 16000
audio_data = np.zeros(16000, dtype=np.float32)  # 1 second of silence

# Transcribe
text = model.stt((sample_rate, audio_data))
print(f"Transcription: {text}")

Loading Audio Files

If you've installed with the audio extras:

from fastrtc_whisper_cpp import get_stt_model, load_audio

# Load model
model = get_stt_model()

# Load audio file (automatically resamples to 16kHz)
audio = load_audio("path/to/audio.wav")

# Transcribe
text = model.stt(audio)
print(f"Transcription: {text}")

Using with FastRTC

from fastrtc_whisper_cpp import get_stt_model

# Create the model
whisper_model = get_stt_model()

# Use within FastRTC applications
# (Follow FastRTC documentation for integration details)

Available Models

The package supports various Whisper.cpp models with different sizes and quantization levels:

• English-only models (faster, smaller):

  • tiny.en, tiny.en-q5_1, tiny.en-q8_0
  • base.en, base.en-q5_1, base.en-q8_0
  • small.en, small.en-q5_1, small.en-q8_0
  • medium.en, medium.en-q5_0, medium.en-q8_0

• Multilingual models:

  • tiny, tiny-q5_1, tiny-q8_0
  • base, base-q5_1, base-q8_0
  • small, small-q5_1, small-q8_0
  • medium, medium-q5_0, medium-q8_0
  • large-v1
  • large-v2, large-v2-q5_0, large-v2-q8_0
  • large-v3, large-v3-q5_0
  • large-v3-turbo, large-v3-turbo-q5_0, large-v3-turbo-q8_0

Example:

from fastrtc_whisper_cpp import get_stt_model

# Choose a specific model
model = get_stt_model("medium.en-q8_0")

Advanced Configuration

You can configure the model with specific parameters:

from fastrtc_whisper_cpp import WhisperCppSTT

# Configure with specific model and models directory
model = WhisperCppSTT(
    model="medium.en",
    models_dir="/path/to/models"  # Optional custom models directory
)

Requirements

• Python 3.10+
• numpy
• pywhispercpp
• librosa (optional, for audio file loading)
• click (for CLI features)

Development

Clone the repository and install in development mode:

git clone https://github.com/mahimairaja/fastrtc-whisper-cpp.git
cd fastrtc-whisper-cpp
pip install -e ".[dev,audio]"

License

MIT