"Bridging the gap between privacy and power in mobile AI—where enterprise capabilities meet consumer devices."
A privacy-focused mobile ecosystem for AI inference. Run offline models locally or connect to 100+ cloud models. No subscriptions. No data harvesting. Complete control.
ToolNeuron is an open-source Android application that delivers enterprise-grade AI capabilities to smartphones through a hybrid architecture prioritizing user control and data sovereignty.
Traditional mobile AI applications force users to choose between two unsatisfactory options: limited offline functionality or cloud-dependent platforms that compromise privacy. ToolNeuron resolves this false dichotomy.
Three Operating Modes:
- Privacy Mode — Execute quantized GGUF models entirely on-device using
llama.cpp. Your data never leaves your phone. - Power Mode — Connect to 100+ premium models (GPT-4, Claude 3.5, Llama 3, Gemini) via OpenRouter for complex tasks.
- Hybrid Intelligence — Seamlessly switch between modes mid-conversation while preserving full context.
Local Execution
Native support for GGUF model formats with zero-latency inference. All processing occurs on-device with no network dependencies.
Cloud Orchestration
Unified API integration through OpenRouter provides access to 100+ state-of-the-art models without vendor lock-in.
Intelligent Streaming
Real-time token generation with context-aware memory management ensures smooth performance across both local and cloud deployments.
Powered by Sherpa-ONNX, ToolNeuron includes 11 professional-grade voices (5 American Female, 2 American Male, 2 British Female, 2 British Male) that run entirely on CPU/NPU with zero cloud dependencies and near-instantaneous synthesis.
Available Now:
- Web Search — Real-time information retrieval
- Web Scraper — Extract and inject context from any URL
- DataHub — Mount JSON/text datasets to dynamically enhance model knowledge
In Development:
- Code execution environments
- Image processing pipelines
- Document analysis tools
- Conversation Persistence — Full chat history retention with efficient storage
- Dynamic Datasets — Attach custom knowledge bases without model retraining
- Context Preservation — Switch models mid-conversation without losing thread continuity
| Chat Interface Multi-modal conversations |
Model Hub 100+ models at your fingertips |
Code Canvas Syntax highlighting & export |
DataHub Dynamic context injection |
 |
 |
 |
 |
| Feature | ToolNeuron | Traditional AI Apps |
|---|---|---|
| Offline GGUF Models | ✓ Native Support | ✗ Cloud Only |
| Model Freedom | ✓ 100+ Options | ✗ Vendor Lock-in |
| Content Policy | ✓ Uncensored | ✗ Heavy Filtering |
| Privacy Architecture | ✓ Local-First | ✗ Server Logging |
| Offline TTS | ✓ 11 Premium Voices | ✗ Cloud Dependencies |
| Pricing Model | ✓ Free (BYOK/Model) | ✗ $20-60/month |
| Source Availability | ✓ Apache 2.0 | ✗ Proprietary |
Visit ToolNeuron on APKPure for the latest stable release with automatic update notifications.
Download the latest release from GitHub Releases and install ToolNeuron-Beta-5.1.apk on Android 8.0+ devices.
# Clone repository
git clone https://github.com/Siddhesh2377/NeuroVerse.git
cd NeuroVerse
# Open in Android Studio (Ladybug or newer)
# Sync Gradle dependencies
./gradlew assembleDebug
# Install on connected device
./gradlew installDebug- Download a GGUF model from HuggingFace (recommended:
Llama-3-8B-Q4_K_M.gguf) - Navigate to Settings → Local Models → Import Model
- Select your downloaded GGUF file
- Begin completely offline inference
- Generate an API key at OpenRouter.ai
- Navigate to Settings → API Configuration
- Enter your OpenRouter API key
- Access 100+ models immediately
- Operating System: Android 8.0+ (API 26)
- RAM: 4GB
- Storage: 2GB available space
- Use Case: Cloud models only
- Operating System: Android 14+
- RAM: 8GB or greater
- Processor: Snapdragon 8 Gen 1 or equivalent
- Storage: 5GB+ available space (for local models)
- NPU: Optional but significantly improves performance
- Advanced TTS with multi-voice conversation simulation
- Speech-to-Text via offline Whisper/Sherpa implementation
- Comprehensive export system for conversations, code snippets, and DataHub configurations
- Native TFLite and ONNX execution support
- On-device Stable Diffusion (quantized) for image generation
- Vector memory implementation for long-term context retention using embeddings
- Multi-modal model support with vision capabilities (LLaVA, GPT-4V)
- Cross-device synchronization with desktop companion applications (Windows/Linux)
- Community-driven plugin marketplace
ToolNeuron implements modern Android development patterns:
- Language: Kotlin
- UI Framework: Jetpack Compose
- Local Inference:
llama.cpp(C++ with JNI bindings) - TTS Engine: Sherpa-ONNX
- API Layer: Retrofit + OkHttp
- Database: Room (SQLite)
- Async Operations: Kotlin Coroutines + Flow
We welcome contributions from developers, researchers, and AI enthusiasts. To contribute:
- Fork the repository
- Create a feature branch (
git checkout -b feature/AmazingFeature) - Commit your changes with descriptive messages
- Push to your branch (
git push origin feature/AmazingFeature) - Open a Pull Request with detailed description
Priority Areas:
- Bug reports and fixes
- Documentation improvements
- Internationalization (i18n)
- Cross-device testing and optimization
- Feature implementations aligned with roadmap
Distributed under the Apache 2.0 License. See LICENSE for complete terms.
Permissions:
- Commercial use
- Modification and distribution
- Private use
- Patent use
Conditions:
- License and copyright notice required
- State changes documentation required
"If I have seen further, it is by standing on the shoulders of giants." — Isaac Newton
ToolNeuron builds upon exceptional open-source work:
- llama.cpp — Efficient local inference implementation
- Sherpa-ONNX — Premium offline text-to-speech synthesis
- OpenRouter — Unified API gateway for diverse model ecosystems
- Jetpack Compose — Modern declarative UI framework
- Discord Community: Join discussions and get support
- Issue Tracker: Report bugs or request features
- GitHub Discussions: Technical questions and ideas
Built by Siddhesh2377 and the Open Source Community
If you find ToolNeuron valuable, please consider starring the repository.
Report Bug • Request Feature • View Releases • Download on APKPure
