Keyframes without manual screenshots
FrameNotes detects meaningful visual changes, filters repeated frames, and keeps the images that make a video easier to review.
Beta: short-video MVP, local upload recommended
Turn Videos into Actionable Notes.
No manual screenshots. No line-by-line note taking. FrameNotes extracts keyframes, transcript segments, and structured outputs so learning videos become reviewable notes.
Supports 5-minute beta tests, 100 MB local upload, single-task processing.
Link conversion may be affected by platform access restrictions. For reliable results, upload a local MP4, MOV, AVI, or MKV file.
Matched transcript
The caption appears directly under the image, so the reader can understand what was said at that moment without switching back to the video.
Live Demo Preview
One result page, with image and transcript kept together.
The product goal is simple: turn a video into a scrollable reading document. Instead of a folder full of screenshots or a raw transcript, FrameNotes puts the useful visual context and the corresponding words into the same timeline.
- Compared with manual screenshots: less repetitive capture work.
- Compared with raw transcript: easier to recover visual context.
- Compared with video bookmarks: easier to export, review, and share.
Core Features
Built around the real workflow: watch less, review more.
Transcript aligned to the timeline
Cloud ASR providers turn clear audio into readable transcript segments. When timestamps are available, captions are matched under the right image.
Readable notes you can keep
Generate HTML, Markdown, and presentation-style pages for study, research, and review workflows.
Use Cases
For learning, research, and creator workflows.
Course learning
Podcast notes
YouTube video summaries
Meeting recaps
Lecture review
Research clipping
FAQ
Set expectations before users upload a video.
Does FrameNotes support YouTube?
Yes, link-based conversion is supported, but YouTube access can be affected by platform restrictions. Local video upload is the most reliable path for the current beta.
Do you store my videos?
The current app processes uploads in a server job directory. Product analytics should record task metadata only; uploaded source videos are not intended as permanent user content storage.
How long can a video be?
The public MVP is intentionally limited to short videos. We recommend videos under 5 minutes and local files under 100 MB until the long-video worker pipeline is separated.
What can I download?
Current outputs include HTML, Markdown, JSON, and presentation-style HTML. Word and ZIP export are good next-step improvements.
Why can transcription fail?
Transcription depends on the configured ASR provider, API quota, audio clarity, supported language, and provider availability. Failed tasks should show the provider error instead of fake captions.