Gemini 3.8 Live Voice Assistant Sandbox & Response Latency Simulator

Voice Stream Input Gemini 3.8 Live

Voice Query Simulation

Tone Nuance Detector

Natural

Intuitive

Technical

Casual

Latency & Engine Tuning

Round-Trip Latency 140 ms

Audio Stream Rate 28.4 kbps

Voice Synthesis Persona

Barge-In Interruption

Simulate user speaking while audio is mid-sentence.

Real-Time Synthesized Waveform (Web Audio API) Audio Engine Ready

OPUS 28.4 kbps

BUFFER: 0.0 ms

FREQ: 44.1 kHz

TTFT (First Chunk)

140 ms

✓ Ultra-low latency

Audio Packets

0 pkts

0 kb/s avg

Barge-in Cutoff

-- ms

Listening

Active Tone

Natural

98% match

Live Model Response Transcript 0 / 0 tokens rendered

Click "Start Voice Stream" to simulate real-time query handling with 140ms latency...

Streaming Telemetry Log

[00:00:00.000] SYSTEM_INIT: Gemini 3.8 Live telemetry sandbox initialized. Web Audio synthesis online.