VideoLens — YouTube Video Summaries by Astra Devs
Turn YouTube videos and local files into polished, timestamped reports. Use your own OpenAI key or optional managed Pro.
Some features may require paymentSome features may require payment
1 User1 User
Extension Metadata
Screenshots
About this extension
VideoLens turns the video already open in Firefox—or a local video file—into a polished written report in Firefox's native sidebar. Findings cite timestamped evidence, so you can check their source and context.
Use VideoLens for detailed reports, key insights, tutorials, interviews and podcasts, meeting notes, bug reports, UX reviews, product demos, content reviews, and privacy scans.
Features:
• Analyze YouTube, compatible HTML5 video players, and local MP4, WebM, or MOV files
• Timestamped findings grounded in sampled frames and available captions or transcription
• Independent non-speech sound observations for supported Managed local files after accepting the updated processing notice
• Follow-up questions against the analyzed timeline
• Private local report library with search, backup, restore, reopen, and delete controls
• Export self-contained HTML, print-ready PDF, Markdown, JSON, or copied text
• Reports in English, Simplified or Traditional Chinese, Latin American Spanish, Brazilian Portuguese, Japanese, Korean, French, German, Romanian, or Hindi
Two analysis modes:
Private mode uses your own OpenAI API key. The key stays in Firefox's local extension storage, and the selected frames, audio or captions, page title, prompt, and follow-up questions go directly from Firefox to OpenAI. VideoLens does not receive that analysis content.
Optional Pro mode removes API-key setup. Analysis content is sent through VideoLens and Vercel AI Gateway to OpenAI. Visual analysis, final review and speech transcription use OpenAI. Local-file sound analysis additionally sends short audio windows to DeepInfra serving Xiaomi MiMo after accepting the updated notice. A free account includes three managed starter reports; paid plans provide additional managed reports. Saving a completed report to the cloud library is optional and off by default.
Completed reports and follow-up answers are saved locally by default. Optional managed cloud recovery temporarily stores the selected video, completed processing stages and report for 24 hours; permanent cloud-library saving is separate. VideoLens does not sell your data or use analysis content for advertising.
Important limits: Non-speech sound analysis is available for supported Managed local files up to 20 minutes. Browser-tab captions and Private mode do not analyze non-speech sounds. Acoustic observations are provisional and are kept separate from visual and spoken evidence. DRM-protected players such as Netflix and Disney+ cannot be captured. Live streams are unsupported. Cross-origin video protections on some sites may prevent frame capture.
VideoLens is developed in a private source repository. Use the published browser extension; public source downloads and self-host installation are not offered. Earlier versions distributed under the MIT licence retain their included licence.
Support:
https://videolens.io/support
Privacy policy:
https://videolens.io/privacy
OPTIONAL CLOUD RECOVERY
Local processing remains the default. For written managed reports, you may choose cloud recovery for a single scan. This uploads the full selected video (up to 200 MB and 20 minutes), or downloads a public YouTube video on the server. VideoLens keeps the video, completed processing stages and report in private, account-bound recovery storage for 24 hours, then deletes expired data through hourly maintenance. Wait until the upload is secured before closing the browser. Reopen Recent cloud scans with the same account to continue without another report credit. After a server restart, reconnect to resume. Visual analysis and speech transcription use OpenAI; actual sound analysis is a separate optional choice using DeepInfra/Xiaomi MiMo through Vercel AI Gateway. Temporary recovery does not enable permanent cloud-library saving. Private/BYOK scans, lessons, recipes, procedures and unsupported browser tabs keep local processing.
Use VideoLens for detailed reports, key insights, tutorials, interviews and podcasts, meeting notes, bug reports, UX reviews, product demos, content reviews, and privacy scans.
Features:
• Analyze YouTube, compatible HTML5 video players, and local MP4, WebM, or MOV files
• Timestamped findings grounded in sampled frames and available captions or transcription
• Independent non-speech sound observations for supported Managed local files after accepting the updated processing notice
• Follow-up questions against the analyzed timeline
• Private local report library with search, backup, restore, reopen, and delete controls
• Export self-contained HTML, print-ready PDF, Markdown, JSON, or copied text
• Reports in English, Simplified or Traditional Chinese, Latin American Spanish, Brazilian Portuguese, Japanese, Korean, French, German, Romanian, or Hindi
Two analysis modes:
Private mode uses your own OpenAI API key. The key stays in Firefox's local extension storage, and the selected frames, audio or captions, page title, prompt, and follow-up questions go directly from Firefox to OpenAI. VideoLens does not receive that analysis content.
Optional Pro mode removes API-key setup. Analysis content is sent through VideoLens and Vercel AI Gateway to OpenAI. Visual analysis, final review and speech transcription use OpenAI. Local-file sound analysis additionally sends short audio windows to DeepInfra serving Xiaomi MiMo after accepting the updated notice. A free account includes three managed starter reports; paid plans provide additional managed reports. Saving a completed report to the cloud library is optional and off by default.
Completed reports and follow-up answers are saved locally by default. Optional managed cloud recovery temporarily stores the selected video, completed processing stages and report for 24 hours; permanent cloud-library saving is separate. VideoLens does not sell your data or use analysis content for advertising.
Important limits: Non-speech sound analysis is available for supported Managed local files up to 20 minutes. Browser-tab captions and Private mode do not analyze non-speech sounds. Acoustic observations are provisional and are kept separate from visual and spoken evidence. DRM-protected players such as Netflix and Disney+ cannot be captured. Live streams are unsupported. Cross-origin video protections on some sites may prevent frame capture.
VideoLens is developed in a private source repository. Use the published browser extension; public source downloads and self-host installation are not offered. Earlier versions distributed under the MIT licence retain their included licence.
Support:
https://videolens.io/support
Privacy policy:
https://videolens.io/privacy
OPTIONAL CLOUD RECOVERY
Local processing remains the default. For written managed reports, you may choose cloud recovery for a single scan. This uploads the full selected video (up to 200 MB and 20 minutes), or downloads a public YouTube video on the server. VideoLens keeps the video, completed processing stages and report in private, account-bound recovery storage for 24 hours, then deletes expired data through hourly maintenance. Wait until the upload is secured before closing the browser. Reopen Recent cloud scans with the same account to continue without another report credit. After a server restart, reconnect to resume. Visual analysis and speech transcription use OpenAI; actual sound analysis is a separate optional choice using DeepInfra/Xiaomi MiMo through Vercel AI Gateway. Temporary recovery does not enable permanent cloud-library saving. Private/BYOK scans, lessons, recipes, procedures and unsupported browser tabs keep local processing.
Rated 0 by 0 reviewers
Permissions and data
Optional permissions:
- Access your data for api.openai.com
Required data collection, according to the developer:
- Authentication information
- Browsing activity
- Website content
Optional data collection, according to the developer:
- Personally identifying information
- Technical and interaction data
More information
- Add-on Links
- Version
- 0.8.6
- Size
- 248.84 KB
- Last updated
- 17 days ago (Nov 3, 2026)
- Related Categories
- License
- MIT License
- Privacy Policy
- Read the privacy policy for this add-on
- Version History
- Add to collection