v0.2.0 - Multi-Model Routing

🚀 What's New in v0.2.0

Multi-Model Routing

• Smart routing now uses different models based on query complexity
• Simple queries → TinyLlama (fast, <10s)
• Complex queries → Llama 3.2 (better quality)

Improved Classifier

• Added system design keywords detection
• Better scoring for technical queries
• Adjusted thresholds for more accurate routing

Better Error Handling

• Detailed timeout messages
• Improved Ollama integration
• Configurable timeouts for slow machines

v0.1.0 - Initial Release

SmartLLM Router v0.1.0

First release of the Intelligent LLM Cost Optimizer.

Features

• ✅ Query complexity classification (simple/medium/complex)
• ✅ Multi-provider support (OpenAI, Anthropic, Ollama)
• ✅ Semantic caching with similarity matching
• ✅ Cost tracking and analytics
• ✅ Docker deployment ready

Quick Start

pip install -r requirements.txt
uvicorn src.main:app --reload

See README for full documentation.