Jul 15, 2026: AI_
Snowflake Cortex AI_TRANSCRIBE now supports AAC (Advanced Audio Coding) input, both as a standalone audio file and as the audio track inside a video file. Because AAC is the default audio codec for Apple devices, podcasts, and most streaming and video pipelines, you can now transcribe a much wider range of real-world recordings directly in Snowflake, without first converting them to another format.
Key use cases include:
- Transcribe recordings from Apple and mobile devices: iPhone Voice Memos, QuickTime, and many mobile apps capture
voice as AAC (commonly with an
.m4aextension). Transcribe these recordings directly, without an extra conversion step. - Process podcast and media libraries: AAC is the standard format for podcast and streaming audio distribution. Transcribe episodes and media archives to power search, summarization, and content analysis.
- Analyze video soundtracks: AAC is the default audio track in MP4 and MOV video. Transcribe webinars, interviews, and marketing videos whose audio is encoded as AAC.
- Transcribe contact center and call recordings: Many call recording and streaming platforms store captured audio as AAC. Transcribe support calls and combine the output with other Cortex AI Functions for sentiment and quality analysis.
- Simplify audio ingestion pipelines: Ingest AAC files as-is and transcribe them with a single SQL query, eliminating custom transcoding steps before analysis.
For more information, see AI_TRANSCRIBE.