Multimodal Notes × Graph RAG × 9 Output Engines × MCP Agents 100% local deployment. Your data never leaves your machine.
| OmniNote | Traditional Notes | Cloud AI Notes | |
|---|---|---|---|
| Data Privacy | ✅ 100% local, E2E encrypted | ✅ Local storage | ❌ Uploaded to cloud |
| Multimodal | ✅ Text / Image / Audio / Video | ❌ Text only | |
| Knowledge Graph | ✅ Graph RAG triple fusion | ❌ None | ❌ None |
| Structured Output | ✅ 9 engines (reports, mindmaps, PPT…) | ❌ None | |
| AI Agents | ✅ MCP protocol, infinite extensions | ❌ None | ❌ None |
| Hardware | ✅ Runs on CPU-only | ✅ Low | — |
Triple-fusion retrieval: BM25 keywords + Vector embeddings + Knowledge graph, combined with entity extraction and community clustering. Your notes don't just store information — they understand it.
- Adaptive chunking (parent-child hierarchy)
- Multi-KB isolation & management
- ReAct Agent for intelligent multi-step reasoning
Transform knowledge base content into actionable deliverables with one click:
| Engine | Output | Engine | Output |
|---|---|---|---|
| 📊 Research Report | Markdown long-form | 🗺️ Mind Map | Markmap visualization |
| 🃏 Flashcards | Study cards | 📋 Quiz | Self-assessment |
| 📰 Infographic | Visual poster | 📖 Wiki | Knowledge pages |
| 📑 PPT Outline | .pptx presentation | 🎙️ Podcast | Audio (planned) |
| 🎬 Video | Video (planned) |
Built on the MCP (Model Context Protocol) — connect weather, train tickets, Excel, and any third-party tool through a single conversational interface.
Notes are fully decoupled from models. Switch providers with one click:
| Capability | API Standard | Default Model |
|---|---|---|
| Vision-Language | OpenAI Chat | Qwen3.5-4B |
| Embedding | OpenAI Embedding | Qwen3-Embedding-0.6B |
| Reranker | Jina Rerank | Qwen3-Reranker-0.6B |
| Image Generation | Stable Diffusion | Tiny-SD |
| Speech Recognition | FunASR | Paraformer-large |
- 100% Local — No internet required. SQLCipher end-to-end encryption. Your data stays private.
- CPU-Only Inference — Low-bit quantization delivers 50%+ faster inference. Even old hardware runs AI smoothly.
- Flexible Deployment — Single machine / LAN multi-user / Docker containerized.
# Extract
tar -xf omni-note-linux-x64.tar.xz
# Launch
chmod a+x ./OmniNote && ./OmniNote
# Open browser
open http://localhost:7000See 📖 Quick Start Guide for full deployment instructions.
- Arm 2024 AI Innovation Application Contest — Excellence Award
- Intel 2025 AI Innovation Application Contest — Winner Award
- Lenovo Title Sponsor Special Award
- Qualcomm Snapdragon AI Contest — Excellence Award
👤 User Docs (user/)
| Directory | Content |
|---|---|
user/user-guide/ |
Quick Start · Nginx Deploy · PPT Templates |
| 🌍 Website | https://turingevo.com |
| 📝 Blog | https://turingevo.blog.csdn.net |
| 💻 GitHub | https://github.com/turingevo |
| 🤗 Hugging Face | https://huggingface.co/turingevo |
| wmx@turingevo.com |