Speech Playground

Sign in with your @skit.ai account to continue

Speech Bench

Compare TTS and ASR models side by side

On-Prem
Third Party
Model is cold starting...
The GPU server is waking up from idle. This typically takes 1-3 minutes. Auto-retrying...
Waiting: 0s
Total (E2E)
-
TTFB (Client)
-
Server Processing
-
Download
-
Network Overhead
-
Audio Size
-
Audio Duration
-
Audio Duration
-
Processing Time
-
Realtime Factor
-
TTFT
-
Tokens
-
Avg Inter-token
-
P99 Inter-token
-