Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.gradium.ai/llms.txt

Use this file to discover all available pages before exploring further.

Telephony providers commonly send 8 kHz mono audio encoded as mu-law or A-law. Gradium supports those formats directly for STT and can produce telephony-friendly TTS output.
Use caseGradium formatNotes
Incoming PSTN audio to STTulaw_8000 or alaw_8000Use the encoding your telephony provider sends.
Incoming linear PCM to STTpcm_8000 or pcm_1600016-bit signed little-endian mono PCM.
Outgoing audio to PSTNulaw_8000 or alaw_8000Avoid resampling in your telephony bridge when possible.
Higher quality app audiopcm_24000, pcm_48000, wav, or opusUse outside PSTN constraints.

​
STT WebSocket Setup

{
  "type": "setup",
  "model_name": "default",
  "input_format": "ulaw_8000",
  "json_config": {"language": "en"}
}
Then send base64-encoded audio payloads: 
{"type": "audio", "audio": "base64_encoded_ulaw_chunk"}

​
TTS WebSocket Setup

{
  "type": "setup",
  "voice_id": "YTpq7expH9539ERJ",
  "output_format": "ulaw_8000"
}
The audio messages contain base64-encoded mu-law chunks that can be forwarded to a telephony media stream.

​
Chunk Size Guidance

Many telephony media streams send 20 ms frames. Gradium accepts small chunks, but batching to roughly 80 ms can reduce message overhead while keeping latency low:
Format20 ms payload80 ms payload
ulaw_8000 / alaw_8000160 bytes640 bytes
pcm_8000320 bytes1280 bytes
pcm_16000640 bytes2560 bytes

​
Bridge Checklist

  • Preserve the provider’s stream/session ID in your own logs.
  • Use one Gradium STT session per caller audio stream.
  • Set input_format to the actual bytes you forward.
  • For TTS, request an output format your provider can play directly.
  • Use STT step messages or your provider’s VAD to decide when to send the user’s turn to an agent.
  • Close the WebSocket and start a new session if the call leg changes format.

Speech-to-Text WebSocket

Real-time transcription with VAD and flush.

Text-to-Speech WebSocket

Stream generated audio back to your telephony bridge.