coqui-ai/STT

Health v2 · maintenance only

Adoption not part of the score

2606 stars · 296 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

Coqui STT is an open-source deep learning toolkit for training and deploying speech-to-text models, built on TensorFlow with bindings for multiple programming languages. It offers pre-trained models, streaming and real-time inference, and multi-GPU training, though the project is no longer actively maintained.

Use cases

• transcribe audio recordings to text
• build a voice-controlled app with speech recognition
• train a custom speech-to-text model on my own dataset
• run offline speech recognition on device
• stream live microphone audio to text in real time
• get multiple transcript candidates with confidence scores

When to choose

• you need a self-contained, offline speech-to-text engine with small footprint
• you want to fine-tune or train an ASR model with multi-GPU support
• you need language bindings beyond Python for embedding STT in other apps

When to avoid

• you want actively maintained software or up-to-date model accuracy
• you need state-of-the-art transcription quality, where Whisper-class models are better
• you rely on the online Model Zoo, which is no longer hosted

Facets

library · maturity abandoned

speech-recognition machine-learning deep-learning llm-inference speech-processing machine-learning deep-learning cross-platform python cpp cli speech-to-text asr tensorflow voice-recognition streaming-inference pretrained-models coqui natural-language-processing

Member repositories

For agents

markdown · JSON · MCP: product_card(name="coqui-ai/STT")