videodb

Use the VideoDB SDK: upload the video, call video.index_spoken_words(force=True), then video.search(query) to get matching shots with timestamps. Wrap the search in try/except because it raises InvalidRequestError when no results are found.

VideoDB burns subtitles server-side. Index spoken words on the video first, then call video.add_subtitle() to get a stream URL with subtitles embedded. You can customize styling with SubtitleStyle parameters like font size and color.

No, desktop capture currently supports macOS only. The CaptureClient records screen, microphone, and system audio on macOS, streaming events over WebSocket. For other platforms, use RTSP/RTMP live stream ingestion instead.

This exception is raised whenever a search returns zero matches; it is expected behavior, not a failure. Catch InvalidRequestError, check for the "No results found" message, and treat it as an empty result set.

The timeline editor does not support transitions, speed changes, crop/zoom, color grading, volume mixing, picture-in-picture video, or keyframe animation. For those operations you must fall back to local tools like ffmpeg or moviepy.

Call video.reframe(start=0, end=60, target="vertical") for a 9:16 conversion, preferably on a short segment since reframe is slow. For full-length videos, pass a callback_url for asynchronous processing to avoid timeouts.