Audio analysis
Gemini audio analysis
Transcribe a recording, summarise what was said and locate sections to check. Choose the Gemini app for a first try or the API for a repeatable workflow.
Try it. Then check it. ↗Task guide · Audio analysis
Transcribe and analyse recordings
Ask a model to transcribe, summarise or identify sections of recorded audio.
What AI can take over
Ask a model to transcribe, summarise or identify sections of recorded audio.
Your remaining responsibilities
- Listen at the returned timestamps and check names and numbers.
- Keep the original recording for comparison.
- Use separate editing software if audio needs to be cut.
Tools for the job
1 toolStart small
Give it something real to do.
01
Transcribe a short clip
Include a few known names or numbers as a check.
02
Ask for timestamps
Request the start and end of the sections you care about.
03
Inspect the boundaries
Do not assume a suggested timestamp is an accurate edit point.