Process your first recording

Outcome

You have a session with a transcript and a generated summary that you can review.

A non-sensitive audio or video file you are permitted to process.
An installed Whisper model. Synthezia offers to download one when it is missing.
For a Local Mode summary, an installed local Ollama model.

Select Upload Audio in the title bar or press Cmd+O.
Choose a local file from the picker. Synthezia lists common audio and video extensions, but an allowed extension is not a tested codec guarantee.
If a model dialog appears, choose Download and continue and wait for it to complete.
A session opens and moves through conversion and transcription. Do not close the app while this first job is running.
Review the generated transcript, then inspect the summary. For a Local Mode summary, Synthezia uses the local Ollama runtime; an External API changes that boundary.
Open the session information drawer to add accurate context before regenerating a summary when needed.

The session status reaches Completed after a summary is stored. You can then ask questions about it or export its structured summary.

Conversion failed: retain the source file and error message; a listed extension does not prove the selected codec can be decoded.
No speech was detected: try a recording with clearer speech or choose a more suitable Whisper model.
Summary is not useful: verify the transcript and session context, then regenerate with a suitable summary mode.