Skip to main content

Video Transcription Demo

vlm-1 returns timed segments for a video.

Analyzing Video Content

The sample is the full 1 hour 40 minute Google Cloud Next 25 Opening Keynote.

Understanding the Output

Example Video Transcription

Key Features

  • Long-form. Videos up to 4+ hours, with automatic segmentation.
  • Visual analysis. Scene descriptions, object recognition, and text extraction.
  • Batch. batch=True handles large video collections.

Try Video and Audio to JSON

Use Video -> JSON or Audio -> JSON on VLM Run. Sign up on the platform.