Libraries & frameworks
speech-to-text-benchmark
Picovoice/speech-to-text-benchmark
A minimalist, extensible Python framework for benchmarking speech-to-text engines, comparing accuracy, latency, memory and compute across local and cloud providers on standard datasets.
Topics
Related projects
vosk-api15.2kAn offline speech recognition toolkit with small models and bindings for many languages, built for apps and devices.
willow-inference-server511Willow Inference Server is a self-hosted, CUDA-optimized inference server exposing Whisper speech recognition, text-to-speech and LLM inference over REST, WebRTC and WebSocket APIs.
faster-whisper26.7kA fast Whisper transcription library built on CTranslate2, with quantization, batching, VAD filtering, and word timestamps.
onnx21.7kAn open standard and Python package for AI model interoperability, defining a common graph format across frameworks.
speechbrain11.9kA PyTorch speech toolkit with recipes and pretrained models for recognition, enhancement, separation, and spoken language tasks.
PySyft10.1kA Python library for privacy-preserving data science letting scientists submit computations to run on private data without it leaving owners.