Speech Transcription API

POST https://api.sawalni.com/v1/speech/transcribe

WS wss://api.sawalni.com/v1/speech/transcribe/stream

curl -X POST https://api.sawalni.com/v1/speech/transcribe \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -F "[email protected]" \
  -F "language=ar" \
  -F "model=daktilo"

import requests

response = requests.post(
    "https://api.sawalni.com/v1/speech/transcribe",
    headers={"Authorization": "Bearer YOUR_API_KEY"},
    files={"audio": ("recording.wav", open("recording.wav", "rb"), "audio/wav")},
    data={"language": "ar", "model": "daktilo"},
)

result = response.json()
print(result["text"])
print(f"Duration: {result['duration']}s")

const formData = new FormData();
formData.append('audio', audioFile);
formData.append('language', 'ar');
formData.append('model', 'daktilo');

const response = await fetch('https://api.sawalni.com/v1/speech/transcribe', {
  method: 'POST',
  headers: { 'Authorization': 'Bearer YOUR_API_KEY' },
  body: formData,
});

const result = await response.json();
console.log(result.text);
console.log('Duration:', result.duration, 's');

{
  "type": "start",
  "model": "daktilo",
  "language": "ar",
  "sample_rate": 16000,
  "channels": 1,
  "audio_format": "pcm_s16le"
}

const protocol = 'sawalni.transcribe.v1';
const keyProtocol = 'sawalni-key.' + btoa('YOUR_API_KEY')
  .replace(/\+/g, '-')
  .replace(/\//g, '_')
  .replace(/=+$/, '');

const ws = new WebSocket(
  'wss://api.sawalni.com/v1/speech/transcribe/stream',
  [protocol, keyProtocol]
);

ws.onopen = () => {
  ws.send(JSON.stringify({
    type: 'start',
    model: 'daktilo',
    language: 'ar',
    sample_rate: 16000,
    channels: 1,
    audio_format: 'pcm_s16le',
  }));
};

ws.onmessage = (event) => {
  const message = JSON.parse(event.data);
  if (message.type === 'partial') console.log('partial:', message.text);
  if (message.type === 'sync') console.log('sync:', message.text);
  if (message.type === 'final') console.log('final:', message.text_delta);
  if (message.type === 'done') console.log('done:', message.text);
};

import asyncio
import json
import wave
import websockets

async def main():
    async with websockets.connect(
        "wss://api.sawalni.com/v1/speech/transcribe/stream",
        extra_headers={"Authorization": "Bearer YOUR_API_KEY"},
        subprotocols=["sawalni.transcribe.v1"],
    ) as ws:
        await ws.send(json.dumps({
            "type": "start",
            "model": "daktilo",
            "language": "ar",
            "sample_rate": 16000,
            "channels": 1,
            "audio_format": "pcm_s16le",
        }))
        with wave.open("recording.wav", "rb") as wav:
            assert wav.getframerate() == 16000 and wav.getnchannels() == 1
            while chunk := wav.readframes(3200):
                await ws.send(chunk)
        await ws.send(json.dumps({"type": "stop"}))
        async for message in ws:
            print(message)

asyncio.run(main())

{
  "text": "مرحبا بكم في هذا التسجيل الصوتي",
  "language": "ar",
  "model": "daktilo",
  "duration": 5.76,
  "usage": {
    "audio_bytes": 184320,
    "audio_seconds": 5.76,
    "processing_ms": 1230.5
  }
}