Description
Welcome to Llamazing, the most advanced app for interacting with local artificial intelligence models. Designed for demanding users seeking control and functionality, Llamazing redefines quality standards in mobile AI.
Llamazing emerges as a revolution: a comprehensive and sophisticated platform with an elegant, intuitive design.
Exclusive Features:
Multiple Simultaneous Servers: Manage up to two Ollama servers, optimized for specific tasks with intelligent load balancing. Secure remote access via Tailscale ensures global availability.
Task-Specific Models: Optimize AI for conversations, web searches, and image analysis, ensuring precision and optimized processing times. The app handles automatic switching.
Advanced PDF Analysis: Transform PDFs into interactive knowledge sources with intelligent indexing and local processing, ensuring privacy. Supports long documents and context extraction.
Smart Image Analysis: Unleash the power of computer vision with advanced models, detailed analysis, optimized processing, and smart compression.
Web Search: Keep your AI updated with real time information, intelligent analysis, and synthesis of relevant information with verifiable sources.
Fully Adjustable Parameters: Total control with real-time adjustments (temperature, top-P, context length), democratizing access to advanced features.
Unique Differentials:
Premium Experience: Visceral and enjoyable design, smooth animations, light/dark themes, and smart haptic feedback, optimized for iOS.
Privacy and Security: Full control over your data. All conversations, documents, and analyses remain on your device or servers you control. No data collection.
Advanced Features: Smart conversation history, contextual suggestions, email export, multilingual support (Portuguese/English), and a personal assistant.
Cutting-Edge Technology: Real-time response streaming, parallel processing, smart memory management, and intelligent auto-scroll for a fluid and intuitive experience.
Llamazing offers a powerful, secure, and fully personalized local AI experience, raising the standard of what’s possible on mobile devices.
Nouveautés (v1.7)
1. Native Tool Calling – Complete replacement for the Decision Assistant. The AI now directly calls tools (web search, image generation, MCP) in a single request, eliminating the extra LLM call. This results in significantly faster responses, reduced resource consumption, and a more fluid experience.
2. Clarification Tool – New intelligent tool that allows the AI to ask for clarification when faced with ambiguous requests. The AI presents multiple-choice options, and you simply tap your preferred choice to continue the conversation seamlessly.
3. Mermaid Diagrams – Full support for rendering flowcharts, sequence diagrams, class diagrams, and other Mermaid visualizations directly in chat. Diagrams are rendered with automatic error correction and can be exported as PNG images.
4. Custom STT Models – Added support for offline speech recognition using Sherpa-ONNX. Download and use Whisper, Moonshine, SenseVoice, and Paraformer models for high-quality transcription without internet dependency.
5. Multilingual Kokoro TTS – Compatibility with Kokoro-ONNX multilingual models for natural text-to-speech in multiple languages with improved prosody and voice quality.
6. Voice Chat Performance – Significant optimizations in voice chat processing, including memory management improvements and more efficient model loading/unloading to prevent crashes during extended sessions.
7. Bug Fixes – Multiple stability improvements and interface refinements across the entire application.