v2.1.0: QWEDLocal - Client-Side Verification
🚀 QWEDLocal: Client-Side Verification
The biggest update in QWED history! Run verification directly in your code—no backend server needed.
⭐ Major Features
QWEDLocal Class - Client-side verification
- Works with ANY LLM (OpenAI, Anthropic, Gemini, Ollama)
- $0 cost with local models (Ollama)
- 100% privacy - Data never leaves your machine
- Smart caching - 50-80% cost savings
3 Verification Engines
- ✅ Math (SymPy) - Symbolic validation
- ✅ Logic (Z3) - SAT solving ⭐ NEW
- ✅ Code (AST) - Security analysis ⭐ NEW
Beautiful CLI ⭐ NEW
qwed verify "What is 2+2?"
qwed interactive
📦 Quick Start
from qwed_sdk import QWEDLocal
client = QWEDLocal(base_url="http://localhost:11434/v1", model="llama3")
result = client.verify_math("What is 2+2?")
💰 Cost: $0 (Ollama) to $100/mo (GPT-4) + 50-80% cache savings
📖 Docs: QWED_LOCAL.md | CLI.md | RELEASE_NOTES_v2.1.0.md
⭐ If this saves you time, star the repo!