sttai 0.1.0

Official Python SDK for STT.ai - Speech-to-Text API

STT.ai Python SDK

Official Python client for the STT.ai Speech-to-Text API.

Transcribe audio and video files with state-of-the-art Whisper models, speaker diarization, real-time streaming, and AI-powered summarization.

Installation

pip install sttai

Quick Start

from sttai import STTClient

client = STTClient("your-api-key")

# Transcribe a file
result = client.transcribe("meeting.mp3")
print(result["text"])

# Transcribe with speaker diarization
result = client.transcribe("interview.wav", diarize=True, speakers=2)
for segment in result["segments"]:
    print(f"[{segment['speaker']}] {segment['text']}")

Authentication

Get your API key from stt.ai/account.

Pass it directly or set the STT_API_KEY environment variable:

export STT_API_KEY="your-api-key"

client = STTClient()  # reads from STT_API_KEY

API Reference

Transcribe a File

result = client.transcribe(
    "audio.mp3",
    model="large-v3-turbo",  # Model to use
    language="auto",          # Language code or "auto"
    diarize=True,             # Enable speaker diarization
    speakers=0,               # Number of speakers (0 = auto-detect)
    response_format="json",   # "json", "text", "srt", "vtt", "verbose_json"
)

Transcribe from URL

result = client.transcribe_url(
    "https://example.com/podcast.mp3",
    model="large-v3-turbo",
    language="en",
)

Summarize Text

summary = client.summarize(
    result["text"],
    style="brief",  # "brief", "detailed", "bullets", "action_items"
)
print(summary["summary"])

List Available Models

models = client.models()
for model in models["models"]:
    print(f"{model['id']}: {model['description']}")

List Supported Languages

languages = client.languages()
for lang in languages["languages"]:
    print(f"{lang['code']}: {lang['name']}")

Health Check

status = client.health()
print(status)

Real-Time Streaming

Stream audio for live transcription over WebSocket:

def on_transcript(data):
    print(data["text"], end="\r")

session = client.stream(on_transcript, model="large-v3-turbo", language="en")

# Send audio chunks (PCM 16-bit, 16kHz, mono)
with open("audio.raw", "rb") as f:
    while chunk := f.read(4096):
        session.send(chunk)

# Get the final result
result = session.finish()
print("\nFinal:", result["text"])

Error Handling

The SDK raises specific exceptions for different error types:

from sttai import STTClient, AuthError, RateLimitError, CreditError, STTError

client = STTClient("your-api-key")

try:
    result = client.transcribe("audio.mp3")
except AuthError:
    print("Invalid API key")
except CreditError:
    print("Insufficient credits - top up at stt.ai/pricing")
except RateLimitError as e:
    print(f"Rate limited. Retry after {e.retry_after}s")
except STTError as e:
    print(f"API error ({e.status_code}): {e.message}")

Configuration

client = STTClient(
    api_key="your-api-key",
    base_url="https://api.stt.ai",  # Custom API endpoint
    timeout=300,                      # Request timeout in seconds
)

Requirements

• Python 3.8+
• requests
• websocket-client (for streaming)

License

MIT - see LICENSE for details.