Voxium AI
Processing Time
-
Average Latency
-
Words
-
Settings
VAD Threshold
0.3 (More Sensitive)
0.5 (Default)
0.7 (Less Sensitive)
Silence Threshold:
100ms (Fastest)
200ms (~0.1s)
500ms (~0.2s)
1s (~0.3s) (More accurate)
Batch Size:
4
8 (Default)
16
32
Beam Size:
1 (Faster)
2
4
8 (More Accurate)
Language:
English
Hindi
German
Spanish
Italian
Input Format:
Base64 (PCM)
PCM (raw 16-bit)
μ-law (8-bit)
Sample Rate:
8 kHz
16 kHz (Default)
44.1 kHz
48 kHz
Save Settings