Voice WebSocket Tester

Test the complete STT → LLM → TTS pipeline using high‑fidelity file uploads. Monitor latency per stage, performance grade, and export structured session data.
Debug the Simple WebSocket Voice Pipeline:
Upload audio files (WebM, WAV, MP3, PCM) to test the complete STT → LLM → TTS pipeline via /api/ws/simple-voice-chat endpoint with debug mode enabled. Monitor each stage with detailed metrics and logs.

Configuration

WebM, WAV, MP3, PCM
WSS api.gmai.sa - Simple Voice Chat Secure WebSocket connection
Disconnected

Pipeline Dashboard

Performance: -
Total Time -
Session ID -
Bottleneck -
Stage Status Duration Progress Details
File Upload
Ready -
Waiting for file
Speech-to-Text
Waiting -
Pending upload
Language Model
Waiting -
Pending STT
Text-to-Speech
Waiting -
Pending LLM

Response & Playback

No conversation yet. Upload an audio file to start testing.

TTS Audio: -

Debug Analytics

Processing Time
-
Confidence Score
-
Transcribed Text: -
Language Detected: -
Model Used: -
Audio Duration: -
Audio Format: -
Processing Time
-
Token Usage
- -
Response Text: -
Model Used: -
Temperature: -
Processing Time
-
Audio Quality
Audio Length: -
Voice Used: -
Output Format: -
Text Length: -
Speech Rate: -
00:00:00 INFO SYSTEM Debug interface ready. Upload an audio file to begin testing.

Export all debug data including session information, events log, timing metrics, and configuration to a JSON file downloaded to your local machine.

Download Format: beautyai-debug-data-YYYY-MM-DD-HH-MM-SS.json
File will be downloaded to your browser's default download folder.
Sessions: 0
Avg Response Time: -
Success Rate: -