An example of transcribing an audio file with Whisper and loading its transcription data into a sqlite3 database for easy searching.

transcribe_to_sqlite.py

import whisper
import sqlite3

# Load the model. Pick from tiny, base, small, medium, large.
model = whisper.load_model("base") 

# Transcribe the local file 'podcast.mp3'. Change as necessary
transcription = model.transcribe("podcast.mp3")

# Create the database and table
con = sqlite3.connect("transcription.db")
cur = con.cursor()
cur.execute("CREATE TABLE segments(id, seek, start, end, text)")

# Prepare the segment dicts for bulk insertion
segments = []
for s in transcription['segments']:
  segments.append((s['id'], s['seek'], s['start'], s['end'], s['text']))
  
# Insert the segment tuples into the table
cur.executemany("INSERT INTO segments VALUES(?, ?, ?, ?, ?)", segments)
con.commit()