Scan the visual story
Sample real video frames and find likely boundaries so you can jump quickly between sections.
PIXELOKAY · INTELLIGENT VIDEO TOOLS FLAGSHIP V3
Video Content Analyzer
Turn a long video into a navigable storyboard. Locate visual changes, describe sampled frames with optional on-device AI, review objects and activities, and export searchable timecodes.
◈ Scene timeline✦ Optional local AI⌕ Searchable moments⇩ Creator exports
01 / WORKSPACE
AI Video Content Analyzer
Sample frames · Find visual cuts · Annotate scenes
Video length—Source duration
Scenes0Visual boundaries
Described0Draft or reviewed
Search results0Matched scenes
05 / PREVIEW
Your video, in context
Ready to explore your video
Choose a file to generate a timestamped visual index.
Scene Video duration Playhead
06 / VIDEO INTELLIGENCE
Searchable summary
Analyze a video to build its scene index. Only AI-analyzed or manually reviewed frames receive semantic descriptions.
07 / SCENE LIBRARY
Find, review and edit moments
No scenes yet
0 reviewed
Your timestamped storyboard appears here after analysis.
Page 1
08 / SCENE EDITOR
Make every moment accurate
Select a scene above to edit its description, labels, and review status.
Splitting works on saved projects without the original video. The midpoint is suggested; choose another timestamp if needed.
AI descriptions and object labels can be incorrect. Review them before publishing.
MADE FOR VIDEO EDITORS, EDUCATORS & CREATORS
One file. A searchable creative outline.
Describe and verify
Optionally run downloadable local captioning and object models, then edit each label and description.
Export for editing
Deliver timestamps, human-reviewed summaries, searchable annotations and WebVTT notes to your workflow.
Frequently asked questions
Does it truly understand a complete video?
No. Scene changes are approximate, and optional AI works on representative still frames. Actions requiring temporal understanding, speech transcription, object tracking and precise shot boundaries are not established by this analysis.
Are my videos uploaded?
No video is uploaded by PixelOkay. AI mode downloads free public model/runtime assets from external hosts, then runs inference locally. Network requests to those hosts may disclose your IP address; the video frames stay on your device.
Can I work without AI or internet?
Yes. Visual scene segmentation, manual annotations, preview, and exports require no model. You can use the page offline once its static code is available. AI mode needs its runtime and weights downloaded first, and browser caching varies.
What video formats can I load?
The browser must be able to decode and seek the file. MP4 and WebM are common, but codec support varies. Encrypted, unseekable, or unsupported videos cannot be reliably analyzed.