What does this skill do?

The VideoDB Video Intelligence Skill allows you to process, index, and edit video and audio in the cloud via an API, without relying on ffmpeg or moviepy. Import media from files, URLs, or YouTube; build visual and semantic indexes; search for exact moments using timestamps; and compose timelines with clips, subtitles, and overlays to generate publish-ready multimedia content.

Input from any source
Upload a video from your local files, a URL, or YouTube, and get an instant playback link.
Semantic Search by Scenes
It indexes spoken or visual content and searches for specific moments using timestamps and reproducible evidence.
Subtitles and Transcoding
Generate, translate, or embed SRT/VTT subtitles, and transcode videos by changing the resolution or aspect ratio for social media.
Composition of Timelines
Trim, overlay text, images, or audio, and export the result as a composite HLS stream.

Usage examples

📥 Upload and search
Upload this video from the URL https://example.com/video.mp4, index the spoken content, and search for the segments where the product demonstration is mentioned.
🎬 Edit with subtitles
Create a timeline using the 10- to 30-second segment of the video, add a subtitle with the text "Final," and generate the resulting stream.
📱 Reframe for TikTok
Recrop this video to a 9:16 vertical aspect ratio for TikTok using Smart Mode and transcode it to 720p.
🔍 Visual Search
Index the visual scenes in this video and find the moments when a person appears writing on a whiteboard.

Features

Multi-source intake Upload media from local files, URLs, YouTube, or RTSP streams and get instant playback links.
Dual Indexing Index spoken content using automatic transcription or visual scenes using AI-generated descriptions.
Semantic search with timestamps Search for specific moments within the video using timestamps, relevance scores, and replayable clips.
Composition of Timelines Cut, overlay text, images, or audio, and mix tracks to create composite streams without using local tools.
Transcoding and Reframing Change the resolution, aspect ratio, and quality, or crop to portrait orientation for TikTok and Reels using Smart Mode.

Frequently asked questions

No. All processing (trimming, merging, overlaying, transcoding, and transcription) is performed on the VideoDB server. Do not use local tools when VideoDB supports the operation.
Sign up for free at console.videodb.io and set the VIDEO_DB_API_KEY variable in the project's .env file. The skill does not handle the key directly.
Use `force=True` in `index_spoken_words` to force reindexing. If the scene index already exists, extract the `scene_index_id` from the error message using a regular expression.
Yes. The skill supports the generation, translation, and embedding of subtitles in SRT or VTT format directly from the indexed video content.

Related skills

More skills in the same category:

VideoDB Video Intelligence — Process and edit video in the cloud with Claude AI

¿Prefieres escuchar el contenido? Genera la narración de audio con un clic.