by Siddhesh2377
Complete offline AI ecosystem for Android: Chat (GGUF/LLMs), Images (Stable Diffusion 1.5), Voice (TTS/STT), and Knowledge (RAG Data-Packs), zero subscriptions, no data harvesting. Open-source privacy-first AI on your terms.
# Add to your Claude Code skills
git clone https://github.com/Siddhesh2377/ToolNeuronOffline AI assistant for Android. Run LLMs, generate images, search documents — all on-device. No cloud. No subscriptions. No data leaves your phone.
Download APK · Discord · Report Issue
No comments yet. Be the first to share your thoughts!
.tnbackup| | Minimum | Recommended | |---|---------|-------------| | Android | 10 (API 29) | 12+ | | RAM | 6 GB | 8–12 GB | | Storage | 4 GB free | 10 GB free | | CPU | ARM64 or x86_64 | Snapdragon 8 Gen 1+ |
From the in-app Model Store (recommended):
bartowski/Phi-3.5-mini-instruct-GGUF)Or manually:
.gguf file from HuggingFaceSelect your model, wait for it to load, start typing. Responses stream in real-time.
| Use case | Model | Size | |----------|-------|------| | Quick test | Qwen3.5 0.8B Q4_K_M | ~600 MB | | General use | Qwen3.5 4B Q4_K_M | ~2.8 GB | | Power users | Qwen3.5 9B Q4_K_M | ~5.5 GB |
Pick Q4_K_M for a good balance between quality and size. Use Q6_K if your device has the RAM for it.
| Tool | Status | |------|--------| | Upscaling | Ready | | Segmentation (MobileSAM) | Ready | | Depth estimation | Model pending | | Style transfer | Model pending | | LaMa inpainting | Model pending |
Create knowledge bases from:
.neuron RAG filesThe RAG pipeline uses hybrid retrieval: FTS4 BM25 + vector search + Reciprocal Rank Fusion + Maximal Marginal Relevance. Results are injected into the conversation context automatically.
Encrypted RAGs support admin passwords and read-only user access.
7 built-in plugins the LLM can call during conversations:
| Plugin | What it does | |--------|-------------| | Web Search | Search the web and scrape content | | File Manager | List, read, create files | | Calculator | Math expressions and unit conversion | | Notepad | Save and retrieve notes | | Date & Time | Current time, timezone conversion, date math | | System Info | RAM, battery, storage, device details | | Dev Utils | Hash, encode, format, text transforms |
Inspired by Mem0. After conversations, the LLM extracts facts about you and stores them for future context. Deduplication via Jaccard similarity, with a forgetting curve so stale memories decay. You can view, edit, and delete memories from the Memory screen.
On-device TTS via Supertonic (ONNX Runtime). 10 voices (5 female, 5 male), 5 languages (EN, KR, ES, PT, FR). Adjustable speed and quality. Auto-speak option reads responses aloud.
Auto-detects CPU topology (P-cores, E-cores) and recommends thread count, context size, and cache settings. Three modes: Performance, Balanced, Power Saver.
Export everything to an encrypted .tnbackup file (PBKDF2 + AES-256-GCM):
git clone https://github.com/Siddhesh2377/ToolNeuron.git
cd ToolNeuron
# Debug
./gradlew assembleDebug
./gradlew installDebug
# Release
./gradlew assembleRelease
APKs land in app/build/outputs/apk/.
If you hit NDK issues, make sure NDK 26.x is installed via SDK Manager. For memory issues during build, bump the Gradle heap in gradle.properties:
org.gradle.jvmargs=-Xmx4096m
| Layer | Technology | |-------|-----------| | Language | Kotlin, C++ (JNI) | | UI | Jetpack Compose | | Text inference | llama.cpp | | Image inference | LocalDream (SD 1.5) | | TTS | Supertonic (ONNX Runtime) | | Database | Room + UMS (custom binary format) | | Encryption | AES-256-GCM, Android KeyStore | | DI | Dagger Hilt | | Async | Kotlin Coroutines + Flow |
| Module | Purpose |
|--------|---------|
| app | Main Android application |
| ums | Unified Memory System — binary record storage with JNI |
| neuron-packet | Encrypted RAG packet format with access control |
| memory-vault | Legacy encrypted storage (read-only, used for migration) |
| system_encryptor | Native encryption primitives |
| file_ops | Native file operations |
See CONTRIBUTORS.md for the project ecosystem and related repos.
git checkout -b feature/your-featureIf you find a security vulnerability:
Apache License 2.0 — use it, modify it, distribute it. Attribution appreciated.
Built by Siddhesh Sonar