Search Videos by Spoken Words in iOS 18 Photos

Apple's iOS 18 introduces significant smart search upgrades to the Photos app, enabling users to locate specific video clips and precise scenes by searching for spoken dialogue. Powered by on-device machine learning and Apple Intelligence, the system automatically transcribes and indexes audio within video files, allowing natural language queries to surface exact moments where specific words or phrases were uttered.

On-Device Audio Transcription and Indexing

The core technology behind this feature is automatic, on-device audio transcription. Leveraging the Apple Neural Engine, iOS 18 analyzes the audio tracks of recorded videos in your library in the background. The system generates searchable text indices from spoken words without sending your private media to cloud servers, ensuring full user privacy while enabling deep content discoverability.

Natural Language Search Queries

Instead of relying solely on broad metadata such as dates, locations, or recognized faces, the Photos search bar now accepts natural language queries that describe dialogue. Users can search for direct quotes, specific sentences, or key phrases—such as "happy anniversary," "make a wish," or specific advice given during a speech. The search engine matches these phrases against the transcribed text generated from your videos.

Direct Timestamp Scrubbing

When a query matches spoken dialogue, the Photos app does not simply display the video thumbnail; it identifies the exact timestamp where the words were spoken. Tapping on a search result immediately opens the video at that precise point in the timeline, eliminating the need to manually scrub through long recordings to find a particular moment.

Combined Semantic and Audio Context

The smart search engine combines audio analysis with visual and contextual recognition. This allows users to combine visual prompts with spoken elements—for instance, searching for a specific person speaking a specific phrase, or searching for dialogue spoken during a particular trip or event.

How to Find Video Clips by Spoken Words

  1. Open the Photos app on an iPhone running iOS 18.
  2. Tap the Search tab or icon.
  3. Type the specific word, phrase, or sentence spoken in the video into the search bar.
  4. Review the results under the Videos section.
  5. Tap the matching video clip to jump directly to the segment where the phrase occurs.