Complete Streaming Session

Close the WebSocket, end the session with REST, retrieve results, and troubleshoot for Ambient /ws/stream and Dictation /ws/transcribe

After you finish streaming audio, close the WebSocket and complete the session with REST APIs. Click on each tab below to learn more about how to complete the session after streaming for Ambient and Dictation.

Ambient

Prerequisites

• You already streamed audio on Ambient /ws/stream or Dictation /ws/transcribe for an active session.
• A valid Suki Token. See Partner authentication.
• For Ambient: readiness to send RU9G, close the socket, call End, then poll status and retrieve content.

Complete the session after streaming audio

After sending the final RU9G message, close the WebSocket connection and complete the ambient session with REST.

1. Close the WebSocket Connection

   Close the WebSocket connection from the client when you are done sending audio for that segment.

2. End the Session Using REST

   End the session using End ambient session API.

3. Retrieve Results Using REST

   Poll session status and fetch transcript and structured data using these REST APIs:

   • Poll Session status API.
   • Get Transcript API.
   • Get Structured data API.

Final transcripts and notes are not guaranteed to arrive over WebSocket. Treat REST APIs as the source of truth.

Troubleshooting

Use this table when ambient streaming misbehaves. It explains what each error or behavior means, why it occurs, and how to resolve it.

Problem	Why it happens	How to fix it
Server returns JSON parse errors (for example, invalid character or null byte)	The WebSocket received a binary frame, a non-JSON payload, or multiple JSON objects in a single frame.	Send one UTF-8 JSON object per WebSocket text frame. Do not send binary frames or concatenate multiple JSON objects.
Audio is ignored or transcription is incorrect	Audio was not encoded as expected. Common causes include hexadecimal encoding, URL-safe Base64, or sending WAV headers instead of raw PCM audio.	Encode raw LINEAR16 PCM audio using standard Base64 (RFC 4648) before sending it in the data field.
The session never completes	The server never received the end-of-stream marker.	End the stream by sending a final AUDIO message with "data": "RU9G" (Base64 for EOF).
Control messages are ignored	A control action was sent in the data field instead of as an event message.	Send control actions as {"type":"EVENT","event":"<VALUE>"}. Do not place control values in data.
No final note or transcript is available	Closing the WebSocket does not end the ambient session or trigger note generation.	After streaming completes, call the ambient REST API to end the session, then retrieve the generated note and transcript.
The WebSocket disconnects during a long pause	No messages were sent before the keep-alive timeout expired.	While audio is paused, send a KEEP_ALIVE event at least every 5 seconds. Ambient supports pauses of up to 30 minutes when keep-alives are maintained. See the Quickstart.
Unable to reconnect after a disconnect	The session has already moved beyond the CREATED state.	Reconnect to /ws/stream using the same ambient_session_id only while the session status is CREATED. Otherwise, the handshake returns FailedPrecondition.

Dictation

Complete the session after streaming

After you send AUDIO_END and finish reading results from the WebSocket, close the connection and complete the session with REST.

1. Finish Reading WebSocket Frames

   Commit transcript text from inbound frames where is_final is true. Wait for the terminal EOF frame. End Dictation session does not return that text.

2. Close the WebSocket Connection

   Close the WebSocket connection from the client when you are done sending audio on that connection.

3. End the Dictation Session Using REST

   End the transcription session using End Dictation session. The HTTP body is empty. Do not read final_transcript from End.

End the session with REST after you finish reading frames. The End response body is empty. Use live WebSocket is_final frames as the transcript.

Troubleshooting

Use this table when Dictation streaming misbehaves. It explains what each error or behavior means, why it occurs, and how to resolve it.

Problem	Why it happens	How to fix it
Handshake fails with FailedPrecondition (session is not accepting new speech sessions)	Another speech session on that transcription_session_id is still finishing, or you reconnected too soon after EOF. There is no partner status API for READY / IDLE.	After AUDIO_END and EOF, wait about 5 seconds before you open the next WebSocket.
No final transcript is available, or the session is left open	Closing the WebSocket does not end the Dictation session. End REST does not return the transcript.	After you finish reading frames, call End Dictation session to close the session. Keep is_final text from the WebSocket.
401 on handshake or immediate disconnect	The browser Sec-WebSocket-Protocol string is wrong, or the token and session ID are in the wrong order.	Use SukiAmbientAuth,<sdp_suki_token>,<transcription_session_id>, not SukiTranscriptionAuth. Put the token before the session ID. Only the session ID value differs from Ambient (transcription_session_id vs ambient_session_id).
Server returns JSON parse errors (for example, invalid character or null byte)	The WebSocket received a binary frame, a non-JSON payload, or multiple JSON objects in a single frame.	Send one UTF-8 JSON object per WebSocket text frame. Do not send binary frames or concatenate multiple JSON objects.
Audio is ignored or transcription is incorrect	The payload used the ambient data field, non-PCM bytes, hexadecimal or URL-safe Base64, or WAV headers treated as PCM.	Encode raw PCM_S16LE with standard Base64 (RFC 4648) and send it in audioData. Strip the WAV header (for example 44 bytes) or decode to raw PCM before encoding.
The session never completes	The server never received the end-of-audio message.	After the last AUDIO chunk, send {"type":"EVENT","event":"AUDIO_END"}.
Partial UI flicker or empty words	Every inbound frame is treated as committed text, or words is read from partial frames.	Show is_final: false frames as draft text. Use words on is_final: true frames for word-level or speaker-aware display. Do not treat transcript_id as a stable key for the same utterance across partials.

Available cookbooks

Available tutorials

Next steps

Ambient audio streaming API - ambient WebSocket request and response contract.

Stream audio to a Dictation session - Dictation WebSocket handshake and message fields.

Audio capture and streaming FAQs - Codec, chunk size, and reconnect guidance.