DocsMedia

Media

YouTube, Subtitles, and Voice Input

12 min read

On this page

  1. 01YouTube link actions
  2. 02Transcript pipeline
  3. 03ASR configuration
  4. 04How creators use it
  5. 05Use transcripts as source material
  6. 06Check transcript quality
  7. 07Turn subtitles into concrete outputs
  8. 08Keep media keys scoped

YouTube link actions

When you paste a YouTube link into the input, Lumo can offer actions such as summarize, key points, script, research material, subtitles, and translation.

Transcript pipeline

The sidecar checks cache first, then attempts official captions. If no official captions are available and ASR is configured, it falls back to speech recognition.

Example
cache -> official captions -> ASR fallback

ASR configuration

Remote ASR uses the Groq key configured in YouTube settings. Voice input can also run in local Whisper ONNX mode or remote mode depending on your settings.

How creators use it

Use YouTube and subtitles as source material for topic research, script drafting, subtitle proofreading, and Shorts splitting. Keep source links and transcript artifacts with the content project.

Use transcripts as source material

Transcripts are useful for extracting arguments, examples, structure, vocabulary, and timing. They are not a replacement for your own script. Ask Lumo to summarize, compare, or repurpose the material into your own angle rather than copying another creator's wording.

Check transcript quality

Before using subtitles for writing, scan speaker names, timestamps, proper nouns, numbers, and mixed-language sections. Official captions can still contain errors, and ASR fallback can mishear names or technical terms.

Use transcripts as references. Verify important claims before publishing.

Turn subtitles into concrete outputs

Useful follow-up tasks include: produce a structured brief, find 5 short-clip candidates, proofread SRT timing and wording, write a platform-specific summary, extract title ideas, or compare several videos before planning your own piece.

Keep media keys scoped

If your build uses remote transcription or image services, keep those keys in Settings and avoid pasting them into chat. When a key stops working, check Settings first, then retry the transcript or voice action.

Docs feedback

Tell us if this page does not match your current build.

Send feedback
PreviousNext
Lumo

One focused workspace for every step from creating to publishing.

Product

  • Features
  • Use Cases
  • Pricing
  • Download
  • Changelog

Resources

  • Documentation
  • FAQ
  • Support
  • Blog

Company

  • About
  • Contact

Legal

  • Privacy Policy
  • Terms of Service

© 2026 Lumo. All rights reserved.

|

Built for creators. Free, Pro, and Max plans available now.

Download Lumo
Skip to main content
Lumo
FeaturesUse CasesPricingFAQBlog
Download