{"id":50383439,"url":"https://github.com/clay-good/enklayve-ai","last_synced_at":"2026-05-30T13:30:22.434Z","repository":{"id":325707415,"uuid":"1102092376","full_name":"clay-good/enklayve-ai","owner":"clay-good","description":"Enklayve is a free, local, private, and secure personal AI desktop application, built with Tauri and llama.cpp/Qwen, that provides robust document intelligence capabilities using fast embeddings.","archived":false,"fork":false,"pushed_at":"2026-04-14T23:00:57.000Z","size":1815,"stargazers_count":8,"open_issues_count":0,"forks_count":3,"subscribers_count":1,"default_branch":"main","last_synced_at":"2026-05-29T03:20:51.506Z","etag":null,"topics":["document-intelligence","llm","local-ai","personal-ai","privacy","rust","security","tauri"],"latest_commit_sha":null,"homepage":"https://enklayve.com","language":"Rust","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/clay-good.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","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":null,"dco":null,"cla":null}},"created_at":"2025-11-22T20:00:57.000Z","updated_at":"2026-05-29T00:55:48.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/clay-good/enklayve-ai","commit_stats":null,"previous_names":["clay-good/enklayve"],"tags_count":6,"template":false,"template_full_name":null,"purl":"pkg:github/clay-good/enklayve-ai","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/clay-good%2Fenklayve-ai","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/clay-good%2Fenklayve-ai/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/clay-good%2Fenklayve-ai/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/clay-good%2Fenklayve-ai/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/clay-good","download_url":"https://codeload.github.com/clay-good/enklayve-ai/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/clay-good%2Fenklayve-ai/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":33694714,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-05-26T15:22:16.424Z","status":"online","status_checked_at":"2026-05-30T02:00:06.278Z","response_time":92,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"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":["document-intelligence","llm","local-ai","personal-ai","privacy","rust","security","tauri"],"created_at":"2026-05-30T13:30:19.646Z","updated_at":"2026-05-30T13:30:22.424Z","avatar_url":"https://github.com/clay-good.png","language":"Rust","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Enklayve\n\n**100% Local, Always Free, AI-Powered Document Analysis**\n\nEnklayve is a secure, privacy-first desktop application that lets you chat with your documents using powerful AI models running entirely on your computer. No cloud services, no subscriptions, no data leaks.\n\n[![Platform](https://img.shields.io/badge/Platform-macOS%20%7C%20Windows%20%7C%20Linux-blue)](https://github.com/user/enklayve-dev)\n[![License](https://img.shields.io/badge/License-MIT-green)](LICENSE)\n[![Rust](https://img.shields.io/badge/Rust-1.70%2B-orange)](https://www.rust-lang.org/)\n\n## Why Enklayve?\n\n- **100% Local Processing**: Everything runs on your device. No internet required.\n- **Always Free**: No subscriptions, no hidden costs, no premium tiers.\n- **Private by Default**: Your documents never leave your computer.\n- **Powerful Models**: Run models like Qwen 2.5, Llama 3.2, Mistral locally.\n- **GPU Accelerated**: Automatic optimization for Apple Silicon and NVIDIA GPUs.\n\n## Key Features\n\n### Privacy \u0026 Security\n- End-to-end encryption with AES-256-GCM\n- Biometric authentication (Touch ID / Windows Hello / Linux Fingerprint)\n- Zero-knowledge architecture\n- All data stays on your device\n\n### Document Intelligence\n- Upload PDFs, DOCX, TXT, Markdown, Excel files\n- Ask questions about your documents\n- Get answers with source citations\n- Vector search for relevant context\n\n### Performance\n- Hardware-aware model recommendations\n- Automatic GPU acceleration (Metal, CUDA)\n- Intelligent context window optimization\n- Fast embedding generation\n\n### User Experience\n- Simple mode for beginners\n- Advanced mode for power users\n- Conversation history\n- Export conversations with sources\n- Automatic backup and restore\n\n## System Requirements\n\n### Minimum\n- **RAM**: 8GB (can run 3B models)\n- **Storage**: 10GB available space\n- **OS**: macOS 10.15+, Windows 10+, Ubuntu 20.04+\n\n### Recommended\n- **RAM**: 16GB (can run 7B models smoothly)\n- **GPU**: Apple Silicon M1+ or NVIDIA GTX 1060+\n- **Storage**: 50GB available space\n\n### For 14B+ Models\n- **RAM**: 32GB+\n- **GPU**: Apple M2 Pro/Max or NVIDIA RTX 3060+\n- **Storage**: 100GB available space\n\n## Installation\n\n### macOS (Apple Silicon \u0026 Intel)\n\n1. Download the DMG installer from [Releases](https://github.com/user/enklayve-dev/releases)\n2. Open the DMG file\n3. Drag Enklayve to Applications folder\n4. Run the following command to allow unsigned app:\n   ```bash\n   xattr -cr /Applications/enklayve.app\n   ```\n5. Launch Enklayve from Applications\n\n### Windows\n\n1. Download the EXE installer from [Releases](https://github.com/user/enklayve-dev/releases)\n2. Run the installer\n3. Follow the installation wizard\n4. Launch Enklayve from Start Menu\n\n### Linux\n\n1. Download the AppImage from [Releases](https://github.com/user/enklayve-dev/releases)\n2. Make it executable:\n   ```bash\n   chmod +x enklayve-*.AppImage\n   ```\n3. Run the AppImage:\n   ```bash\n   ./enklayve-*.AppImage\n   ```\n\n## Quick Start\n\n1. **First Launch**: Enklayve will detect your hardware and recommend the best model\n2. **Download Model**: Click \"Download Recommended Model\" (one-time download, 1-20GB depending on model)\n3. **Upload Document**: Drag and drop a PDF or document file\n4. **Ask Questions**: Start chatting about your document\n5. **Get Answers**: Receive intelligent responses with source citations\n\n## Supported Document Formats\n\n- PDF (with OCR support for scanned documents)\n- Microsoft Word (.docx)\n- Plain Text (.txt)\n- Markdown (.md)\n- Excel (.xlsx, .xls)\n- CSV (.csv)\n\n## Available Models\n\nEnklayve automatically recommends the best model based on your RAM:\n\n| RAM | Recommended Model | Size | Performance |\n|-----|-------------------|------|-------------|\n| 8GB | Qwen 2.5 3B | 1.9GB | Fast, efficient |\n| 16GB | Qwen 2.5 7B | 4.4GB | Balanced, recommended |\n| 32GB | Qwen 2.5 14B | 8.7GB | Very smart |\n| 64GB+ | Qwen 2.5 32B | 19GB | Maximum intelligence |\n\nAll models support 32K context window and run at 30-60 tokens/sec on modern hardware.\n\n## Building from Source\n\n### Prerequisites\n- Node.js 20+\n- Rust 1.70+\n- Platform-specific tools:\n  - **macOS**: Xcode Command Line Tools\n  - **Windows**: Visual Studio Build Tools\n  - **Linux**: GCC, libwebkit2gtk-4.1-dev\n\n### Build Steps\n\n```bash\n# Clone repository\ngit clone https://github.com/user/enklayve-dev.git\ncd enklayve-dev/enklayve-app\n\n# Install dependencies\nnpm install\n\n# Run in development mode\nnpm run tauri dev\n\n# Build for production\nnpm run tauri build\n```\n\n### Build Outputs\n\n- **macOS**: `src-tauri/target/release/bundle/dmg/`\n- **Windows**: `src-tauri/target/release/bundle/nsis/`\n- **Linux**: `src-tauri/target/release/bundle/appimage/`\n\n## Project Structure\n\n```\nenklayve-dev/\n├── enklayve-app/          # Main Tauri application\n│   ├── src/               # React frontend\n│   │   ├── components/    # UI components\n│   │   ├── App.tsx       # Main app component\n│   │   └── main.tsx      # Entry point\n│   ├── src-tauri/        # Rust backend\n│   │   ├── src/\n│   │   │   ├── commands.rs        # Tauri commands\n│   │   │   ├── inference.rs       # LLM inference\n│   │   │   ├── documents.rs       # Document processing\n│   │   │   ├── embeddings.rs      # Vector embeddings\n│   │   │   ├── hardware.rs        # GPU detection\n│   │   │   ├── encryption.rs      # AES-256-GCM encryption\n│   │   │   ├── biometric.rs       # Touch ID / Windows Hello / fprintd\n│   │   │   ├── conversations.rs   # Chat history\n│   │   │   ├── backup.rs          # Backup/restore\n│   │   │   ├── export.rs          # Export functionality\n│   │   │   └── model_cache.rs     # Model caching\n│   │   └── Cargo.toml    # Rust dependencies\n│   └── package.json      # Node dependencies\n├── .github/\n│   └── workflows/        # CI/CD pipelines\n└── README.md            # This file\n```\n\n## Architecture\n\n### Technology Stack\n- **Frontend**: React 19 + TypeScript + Vite\n- **Backend**: Rust + Tauri 2.0\n- **LLM Engine**: llama.cpp (via llama-cpp-2)\n- **Embeddings**: BGE-Small-EN-v1.5 (fastembed)\n- **Database**: SQLite with encryption\n- **Vector Search**: Custom cosine similarity\n- **Document Processing**: pdf-extract, docx-rs, calamine (Excel), ocrs (OCR)\n\n### GPU Acceleration\n\n#### Apple Silicon (Automatic)\n- Metal acceleration enabled by default\n- Unified memory architecture for efficient GPU offloading\n- Optimized layer distribution based on available RAM\n\n#### NVIDIA GPUs (Windows/Linux)\n- Automatic CUDA detection\n- Dynamic GPU layer allocation\n- Build with: `cargo build --release --features cuda`\n\n## Troubleshooting\n\n### macOS: \"App can't be opened because the developer cannot be verified\"\nRun this command to remove quarantine attribute:\n```bash\nxattr -r -d com.apple.quarantine /Applications/enklayve.app\n```\n\n### Model download is slow\nModels are large (1-20GB). Download times depend on your internet speed. Downloads can be resumed if interrupted.\n\n### Out of memory during inference\nTry a smaller model. Qwen 2.5 3B works well on 8GB RAM systems.\n\n### App crashes on startup\nCheck system requirements. Ensure you have at least 8GB RAM and 10GB free disk space.\n\n### GPU not detected\n- **NVIDIA**: Install latest drivers and CUDA toolkit\n- **Apple**: Metal is automatic on M1/M2/M3/M4 chips\n\n### Linux: Fingerprint authentication not working\nEnsure fprintd is installed and fingerprints are enrolled:\n```bash\n# Install fprintd\nsudo apt install fprintd\n\n# Enroll fingerprint\nfprintd-enroll\n```\n\n## Privacy \u0026 Security\n\nEnklayve is built with privacy as the foundation:\n\n- **No telemetry**: We don't collect any usage data\n- **No analytics**: No tracking, no crash reports (unless you opt-in)\n- **No network calls**: Models run 100% offline\n- **Encrypted storage**: All data encrypted with AES-256-GCM\n- **Secure credentials**: Platform keychain integration (macOS Keychain, Windows Credential Manager, Linux Secret Service)\n- **Biometric unlock**: Touch ID, Windows Hello, and Linux fingerprint (fprintd) support\n- **Open source**: Full transparency, audit the code yourself\n\n## Performance Tips\n\n1. **Use Simple Mode**: Advanced mode loads all features, Simple mode is faster\n2. **Enable Auto-tuning**: Let Enklayve optimize settings for your hardware\n3. **Smaller Context Window**: Reduce context window if experiencing slowdowns\n4. **Close Other Apps**: Free up RAM for better performance\n5. **Use SSD**: Store models on SSD for faster loading\n\n## Contributing\n\nWe welcome contributions! Please see [CONTRIBUTING.md](CONTRIBUTING.md) for guidelines.\n\n## License\n\nMIT License - see [LICENSE](LICENSE) for details.\n\n## Support\n\n- **Issues**: [GitHub Issues](https://github.com/user/enklayve-dev/issues)\n- **Discussions**: [GitHub Discussions](https://github.com/user/enklayve-dev/discussions)\n- **Documentation**: [docs/](docs/)\n\n## Roadmap\n\n- [ ] Image analysis\n- [ ] Voice input/output\n- [ ] Multiple document comparison\n- [ ] Plugin system\n- [ ] Mobile apps (iOS, Android)\n\n## Acknowledgments\n\nBuilt with:\n- [Tauri](https://tauri.app/) - Desktop app framework\n- [llama.cpp](https://github.com/ggerganov/llama.cpp) - LLM inference engine\n- [Qwen](https://huggingface.co/Qwen) - State-of-the-art language models\n- [fastembed](https://github.com/Anush008/fastembed-rs) - Fast embeddings\n\n---\n\nMade with care for privacy and local-first computing.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fclay-good%2Fenklayve-ai","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fclay-good%2Fenklayve-ai","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fclay-good%2Fenklayve-ai/lists"}