Gladia Python SDK

Developer-friendly Python SDK for Gladia batch and real-time transcription.

Features

• Simple GladiaClient(api_key=...) or GladiaClient.from_env()
• Batch file transcription (sync and async)
• Real-time streaming over WebSocket (async) with helpers for mic and files
• Typed models, retries, timeouts, and logging
• CLI for quick demos: gladia transcribe, gladia stream

Installation

pip install gladia
# Optional extras
pip install "gladia[realtime]"     # WebSocket streaming
pip install "gladia[audio]"        # Mic/file streaming helpers (sounddevice, numpy, webrtcvad)

Quickstart

from gladia import GladiaClient

client = GladiaClient.from_env()  # expects GLADIA_API_KEY
result = client.transcribe_file("/path/to/audio.wav")
print(result.text)

CLI

gladia transcribe /path/to/audio.wav --language en --timestamps --output result.json
gladia stream --mic
gladia stream --file /path/to/audio.wav --fast

Configuration

• Environment variables:
  • GLADIA_API_KEY
  • GLADIA_BASE_URL (default: https://api.gladia.io)
  • GLADIA_REALTIME_URL (default: wss://api.gladia.io/realtime)

License

This SDK is provided under a source-available non-commercial license. For commercial licensing, contact Gladia.