{"id":49889739,"url":"https://github.com/sergiomarquezdev/yt-transcriber","last_synced_at":"2026-05-15T20:10:47.753Z","repository":{"id":341045813,"uuid":"1109827115","full_name":"sergiomarquezdev/yt-transcriber","owner":"sergiomarquezdev","description":"🛠️ CLI tool to transcribe YouTube videos using OpenAI Whisper with CUDA acceleration, generate AI summaries (EN/ES) with Gemini, and create LinkedIn/Twitter content. Supports YouTube, Google Drive, and local files.","archived":false,"fork":false,"pushed_at":"2026-05-06T08:33:25.000Z","size":324,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-05-06T10:38:36.877Z","etag":null,"topics":["ai","cli","cuda","gemini","python","transcription","whisper","youtube"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/sergiomarquezdev.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":"AGENTS.md","dco":null,"cla":null}},"created_at":"2025-12-04T10:38:17.000Z","updated_at":"2026-05-06T08:33:31.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/sergiomarquezdev/yt-transcriber","commit_stats":null,"previous_names":["sergiomarquezdev/yt-transcriber"],"tags_count":1,"template":false,"template_full_name":null,"purl":"pkg:github/sergiomarquezdev/yt-transcriber","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sergiomarquezdev%2Fyt-transcriber","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sergiomarquezdev%2Fyt-transcriber/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sergiomarquezdev%2Fyt-transcriber/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sergiomarquezdev%2Fyt-transcriber/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/sergiomarquezdev","download_url":"https://codeload.github.com/sergiomarquezdev/yt-transcriber/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sergiomarquezdev%2Fyt-transcriber/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":33078262,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-05-15T20:05:40.333Z","status":"ssl_error","status_checked_at":"2026-05-15T20:05:38.672Z","response_time":103,"last_error":"SSL_read: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["ai","cli","cuda","gemini","python","transcription","whisper","youtube"],"created_at":"2026-05-15T20:10:46.992Z","updated_at":"2026-05-15T20:10:47.745Z","avatar_url":"https://github.com/sergiomarquezdev.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# YouTube Video Transcriber \u0026 Summarizer\n\n[![Python 3.12+](https://img.shields.io/badge/python-3.12+-blue.svg)](https://www.python.org/downloads/)\n[![License](https://img.shields.io/badge/license-MIT-green.svg)](LICENSE)\n[![PyTorch](https://img.shields.io/badge/PyTorch-CUDA%2012.8-red.svg)](https://pytorch.org/)\n\n**Transform YouTube videos into searchable text, AI summaries, and social media content.**\n\n## Features\n\n- **CUDA-accelerated transcription** with OpenAI Whisper\n- **Multi-language support** with auto-detection\n- **AI-powered summaries** (EN + ES) via Claude CLI:\n  - Executive summary (2-3 sentences)\n  - Key points (5-10 bullets)\n  - Smart timestamps (inferred from content)\n  - Action items (when applicable)\n- **Social Media Post Kits**: Auto-generate LinkedIn posts + Twitter threads\n- **Optional timestamped segments**: emit `_segments.json` sidecar with Whisper `start/end/text`\n- **Optional visual evidence (V1)**: extract midpoint frames per segment for local files\n- **Multiple input sources**: YouTube, Google Drive, local files\n- **Performance optimized**:\n  - Automatic memory cleanup (no memory leaks)\n  - Auto temp file cleanup\n\n## Quick Start\n\n```bash\n# Clone the repository\ngit clone https://github.com/sergiomarquezdev/yt-transcriber.git\ncd yt-transcriber\n\n# Create virtual environment\npython -m venv venv\nsource venv/bin/activate  # Linux/macOS\n.\\venv\\Scripts\\activate   # Windows\n\n# Install dependencies\npip install -e .\n\n# Configure environment\ncp .env.example .env\n# Ensure `claude` CLI is in PATH (requires Claude Max/Pro subscription)\n\n# Transcribe your first video\nyt-transcriber transcribe --url \"https://www.youtube.com/watch?v=VIDEO_ID\"\n```\n\n## Prerequisites\n\n- **Python 3.12+**\n- **FFmpeg** - Required for audio processing\n- **CUDA 12.8** (Optional) - For GPU acceleration\n- **Claude CLI** (Optional) - Required for AI summaries, translations, and post kits. Install from [Claude Code](https://docs.anthropic.com/en/docs/claude-code) and ensure `claude` is in your PATH with an active subscription.\n\n### FFmpeg Installation\n\n**Windows:**\n```powershell\n# Download from https://github.com/BtbN/FFmpeg-Builds/releases\n# Extract to C:\\ffmpeg and add to PATH\n[Environment]::SetEnvironmentVariable(\"Path\", $env:Path + \";C:\\ffmpeg\\bin\", [EnvironmentVariableTarget]::User)\n```\n\n**macOS:**\n```bash\nbrew install ffmpeg\n```\n\n**Linux:**\n```bash\nsudo apt install ffmpeg  # Ubuntu/Debian\n```\n\n## Usage\n\n```bash\n# Only transcription (DEFAULT)\nyt-transcriber transcribe --url \"https://www.youtube.com/watch?v=VIDEO_ID\"\n\n# Transcription + AI summaries (EN + ES)\nyt-transcriber transcribe --url \"URL\" --summarize\n\n# Transcription + summaries + Post Kits (LinkedIn + Twitter)\nyt-transcriber transcribe --url \"URL\" --post-kits\n\n# Force Spanish transcription\nyt-transcriber transcribe --url \"URL\" --language es\n\n# Local file\nyt-transcriber transcribe --url \"path/to/video.mp4\" --summarize\n\n# Emit timestamped segments sidecar JSON\nyt-transcriber transcribe --url \"URL_OR_LOCAL_FILE\" --segments\n\n# Extract visual evidence frames (implies --segments, local files only in V1)\nyt-transcriber transcribe --url \"path/to/video.mp4\" --visual-evidence\n```\n\n### CLI Options\n\n| Option | Short | Description |\n|--------|-------|-------------|\n| `--url` | `-u` | YouTube URL, Google Drive URL, or local file path |\n| `--language` | `-l` | Language code (`en`, `es`) - auto-detect if omitted |\n| `--summarize` | | Generate AI summaries (EN + ES) |\n| `--post-kits` | | Generate LinkedIn + Twitter content (implies --summarize) |\n| `--segments` / `--no-segments` | | Enable/disable `_segments.json` sidecar (CLI override; env fallback when omitted) |\n| `--visual-evidence` / `--no-visual-evidence` | | Enable/disable frame extraction per segment (local files only in V1; visual implies segments) |\n| `--ffmpeg-location` | | Custom FFmpeg path |\n\n## Output\n\nFiles are saved to `output/`:\n\n```\noutput/\n├── transcripts/\n│   ├── {title}_vid_{id}.txt             # Raw transcription (always)\n│   ├── {title}_vid_{id}_segments.json   # Optional: segments sidecar\n│   └── {title}_vid_{id}_frame_{idx}.jpg # Optional: visual evidence frames\n└── summaries/\n    ├── {title}_vid_{id}_summary_EN.md  # English summary\n    ├── {title}_vid_{id}_summary_ES.md  # Spanish summary\n    └── {title}_vid_{id}_post_kits.md   # LinkedIn + Twitter content\n```\n\n### AI Summary Contents\n\n- Executive summary\n- Key points (5-7 bullets)\n- Timestamps (5-8 important moments)\n- Action items\n\n### Post Kits Contents\n\n- **LinkedIn post** (800-1200 chars): Professional hook, insights, CTA, hashtags\n- **Twitter thread** (8-12 tweets): Numbered tweets, max 280 chars each\n\n## Configuration\n\nCreate `.env` from template:\n\n```bash\n# Whisper Model\nWHISPER_MODEL_NAME=base    # tiny, base, small, medium, large\nWHISPER_DEVICE=cuda        # cuda or cpu\n\n# Claude CLI (required for summaries, translations, post kits)\n# Ensure `claude` is in PATH with active subscription (Max/Pro)\n# CLAUDE_CLI_PATH=claude\n# CLAUDE_CLI_TIMEOUT=180\n# DEFAULT_LLM_MODEL=sonnet\n\n# Directories\nTEMP_DOWNLOAD_DIR=temp_files/\nOUTPUT_TRANSCRIPTS_DIR=output/transcripts/\nSUMMARY_OUTPUT_DIR=output/summaries/\n\n# Optional transcript segments sidecar (default off)\nTRANSCRIPT_SEGMENTS_ENABLED=false\n\n# Optional visual evidence extraction (default off, local files only in V1)\nVISUAL_EVIDENCE_ENABLED=false\nVISUAL_EVIDENCE_MIN_SEGMENT_SECONDS=1.0\n\n# Logging\nLOG_LEVEL=INFO\n```\n\n### Model Selection\n\n| Model | Speed | Accuracy | VRAM | Use Case |\n|-------|-------|----------|------|----------|\n| `tiny` | Fast | Low | ~1GB | Quick drafts |\n| `base` | Good | Medium | ~1GB | **Default - Balanced** |\n| `small` | Medium | Good | ~2GB | Better quality |\n| `medium` | Slow | High | ~5GB | High accuracy |\n| `large` | Slowest | Best | ~10GB | Best quality |\n\n## Programmatic Usage\n\n```python\nfrom yt_transcriber.cli import run_transcribe_command\n\n# Only transcription (default)\ntranscript, _, _, _ = run_transcribe_command(url=\"path/to/video.mp4\")\n\n# With summaries\ntranscript, summary_en, summary_es, _ = run_transcribe_command(\n    url=\"https://www.youtube.com/watch?v=VIDEO_ID\",\n    generate_summary=True,\n)\n\n# With post kits (implies summary)\ntranscript, summary_en, summary_es, post_kits = run_transcribe_command(\n    url=\"https://www.youtube.com/watch?v=VIDEO_ID\",\n    generate_post_kits=True,\n)\n```\n\n\n## Performance \u0026 Resource Management\n\nThe application includes automatic optimizations for production use:\n\n**Memory Management:**\n- Whisper model auto-loads/unloads per video (prevents memory leaks)\n- Automatic garbage collection after each transcription\n\n**Resource Cleanup:**\n- Temp files auto-deleted after processing (even on errors)\n- No manual cleanup needed\n\n**FFmpeg:**\n- Timeout protection (5min) prevents hung processes\n\n## Troubleshooting\n\n**FFmpeg not found:**\n```bash\n# Use direct path\nyt-transcriber transcribe --url \"URL\" --ffmpeg-location \"C:\\ffmpeg\\bin\\ffmpeg.exe\"\n```\n\n**CUDA not available:**\n```bash\n# Check installation\npython -c \"import torch; print(torch.cuda.is_available())\"\n\n# Fall back to CPU in .env\nWHISPER_DEVICE=cpu\n```\n\n**Out of memory:**\n```bash\n# Use smaller model in .env\nWHISPER_MODEL_NAME=tiny\n```\n\n## License\n\nMIT License - see [LICENSE](LICENSE)\n\n## Acknowledgments\n\n- [OpenAI Whisper](https://github.com/openai/whisper) - AI transcription\n- [yt-dlp](https://github.com/yt-dlp/yt-dlp) - Video downloading\n- [Claude Code](https://docs.anthropic.com/en/docs/claude-code) - AI summarization \u0026 translation\n\n---\n\nMade with care by [Sergio Marquez](https://github.com/sergiomarquezdev)\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fsergiomarquezdev%2Fyt-transcriber","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fsergiomarquezdev%2Fyt-transcriber","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fsergiomarquezdev%2Fyt-transcriber/lists"}