1
Initial Setup
Settings → Application Settings
Configure your API keys:
- OpenAI API key: For AI-powered content generation (title/description/tags)
- Local LLM (optional): Use LM Studio or Ollama instead for privacy/cost savings
- Whisper settings: Choose model size and device (CPU vs CUDA GPU)
Settings → Video Profiles
Create your first Video Profile:
- Profile name (e.g., "X4_Foundations", "Minecraft_Survival")
- Videos directory path (where your recordings are stored)
- Game logo for thumbnails (optional)
- AI prompts customization (optional - defaults work great!)
Upload Tab → Add YouTube Channel
Authenticate with YouTube:
- Click "Add Channel" button
- Browser opens → Sign in with Google → Grant permissions
- Channel automatically saved for future uploads
- Note: Requires your own OAuth credentials (see YouTube OAuth Setup)
2
Processing Your First Video
-
Dashboard Tab: Select your Video Profile from dropdown
-
Place video: Put
my_first_video.mp4 in the toupload/ folder shown in Dashboard
-
Transcription Tab: Click "Refresh" → Select video → Click "Transcribe Selected"
- Wait for transcription to complete (progress bar shows status)
- Generated file:
my_first_video_transcript.txt
-
Content/Metadata Tab: Select video → Click "Generate Selected"
- AI generates title, description, and tags from transcript
- Review and edit in built-in editor if needed
- Generated file:
my_first_video_yt.txt
-
Thumbnails Tab: Select video → Click "Generate Selected"
- AI selects interesting frame and generates overlay text
- Preview appears instantly on right side
- Generated file:
my_first_video_123_thumb.jpg
-
Upload Tab: Select video (marked "✓ Ready") → Click "Start Upload"
- Video uploads to YouTube with metadata and thumbnail
- On success: Automatically moves to
Done/ folder
🎉 Congratulations! You've published your first video using Loki Studio!
3
Important Files & Directories
Important Files
- client_secret.json - Your OAuth credentials (root directory)
- VideoProfiles/ - Your video profile configuration files
- Assets/Logos/ - Profile logos for thumbnails
- logs/ - Application logs for troubleshooting
Directory Structure
E:/Videos/Gaming/YourGame/
├── toupload/ ← Place new videos here
├── Done/ ← Successfully uploaded videos
├── Failed/ ← Failed uploads (review and retry)
└── game_logo.png ← Logo for thumbnails (optional)
🎙️ Pro Tip: OBS Audio Channel Configuration
Configure your OBS audio channels for optimal transcription and AI content generation!
- Audio Channel 1 (Desktop Audio): Game audio, system sounds, voice lines
- Used for story-driven games (RPGs, narrative games)
- AI will extract dialogue and story context for better descriptions
- Audio Channel 2+ (Microphone): Your commentary and voice
- Primary source for transcription (pure voice, easier to process)
- Higher accuracy for Whisper AI transcription
Why this matters: Channel 1 (game audio) + Channel 2 (mic) gives AI both your commentary AND in-game dialogue. This creates significantly better titles, descriptions, and tags, especially for story-heavy games!
Tips for Success
- ✅ Test with small videos first (5-10 min) to verify setup
- ✅ Use "Dry Run" in Upload tab to preview before actual upload
- ✅ Check logs/ folder if something goes wrong
- ✅ Keep _final naming for edited videos (auto-detected)
- ✅ Monitor API costs via OpenAI dashboard
- ✅ Configure OBS audio channels as described above for best AI results