When you upload content to Eden, it's automatically transcribed, analyzed, and tagged. This is what makes visual search and AI features possible — and it happens without you doing anything.
What Gets Processed
Content Type | Transcription | Frame Analysis | Tagging |
|---|---|---|---|
Uploaded videos | ✓ | ✓ | ✓ |
YouTube videos | ✓ | ✓ | ✓ |
YouTube Shorts | ✓ | ✓ | ✓ |
Instagram Reels | ✓ | ✓ | ✓ |
Twitter/X posts | Text extracted | — | ✓ |
Audio files | ✓ | — | ✓ |
PDFs | Text extracted | — | ✓ |
Images | — | ✓ | ✓ |
How Processing Works
When you add content, Eden works through three stages:
1. Transcription
Eden converts speech to text. Every word spoken in a video or audio file becomes searchable text with timestamps.
2. Frame Analysis
For videos, Eden analyzes each frame to understand what's visually happening — people, objects, scenes, colors, actions, text on screen.
This is what powers searches like "person wearing red jacket" or "screenshot of the dashboard."
3. Segment Creation
Eden groups content into logical segments (scenes) and generates tags and descriptions for each. This makes it easy to navigate long videos and find specific sections.
Viewing Transcripts
To see a transcript:
Open any video in your workspace
Press
Tto open the Transcription panelClick any line to jump to that moment in the video
The transcript is timestamped and synced with playback.
Viewing Segments & Tags
After processing, videos are divided into segments. Each segment has:
Timestamp range — When it starts and ends
Description — What happens in that section
Tags — Keywords describing the content
This makes long videos navigable. Instead of scrubbing through an hour of footage, you can scan the segments to find what you need.
How This Powers Other Features
Visual Search
Because Eden knows what's in every frame and every word spoken, you can search by describing what you're looking for:
"Blue car driving" → Finds frames with blue cars
"When they mention the deadline" → Finds that moment in the transcript
AI Context
When you connect a video to an AI chat on Canvas, the AI has access to the full transcript and frame analysis. It knows everything that was said and shown.
Quick Navigation
Click on any segment or transcript line to jump directly to that point. No more scrubbing through hours of content.
Processing Status
While content is being processed, you'll see an indicator on the item. Processing time depends on content length:
Content Length | Approximate Time |
|---|---|
5-minute video | ~1 minute |
30-minute video | ~3-5 minutes |
1-hour podcast | ~5-10 minutes |
2-hour video | ~10-15 minutes |
You can view the content while it's processing, but search and AI features work best after processing completes.
Tips for Best Results
Clear audio = better transcription
Videos with clear speech and minimal background noise produce more accurate transcripts.
Check important transcripts
For critical content, skim the transcript to catch any errors. AI transcription is very good but not perfect.
Use segments for long content
For videos over 30 minutes, use the segment view to navigate instead of scrolling through a long transcript.
Common Questions
Can I edit transcripts?
Currently, transcripts are read-only.
What languages are supported?
Eden supports transcription in multiple languages. The language is auto-detected.
What if transcription fails?
If an item fails to process, try re-uploading. If it continues to fail, contact support@eden.so with details.
Is my content stored securely?
Yes. All content is encrypted and stored securely. See our privacy policy for details.
Does processing happen even for stored files?
Yes. Every file and link you add is automatically processed, whether you're actively using Eden or not.



