SOURCE-LINKED INTELLIGENCE
Adjustable VAD threshold for Speech to Text
Speech to Text now accepts a vad_threshold parameter (streaming query param and batch multipart field) to tune the voice-activity gate that skips non-speech audio. Lower values transcribe quieter or noisier speech — useful for narrowband telephony — and 0 disables the gate. See the Speech to Text docs.
Read original source ↗ Open in workspace
- recordType
- page-entry
- evidenceStatus
- publisher-reported
- region
- Global
Evidence & attribution
- xAI Grok Release Notes · 2026-07-23T00:00:00.000Z
First collected: 2026-09-23T00:41:11.323Z. This is not the publication date.