Repository navigation
Releases
TMC v2.0 - INTELLIGENT & PREMIUM
TMC v2.0 - Intelligent Multi-Layer Memory System
What Makes v2.0 Different
TMC v2.0 isn't just vector search - it's intelligent memory with 5 layers:
- 📝 Content Layer - Hash-based exact matching
- 🧩 Semantic Layer - Meaning and concept similarity
- 💭 Emotional Layer - Sentiment and tone analysis
- ⏰ Temporal Layer - Time-aware relevance
- 🎯 Importance - Priority-based filtering
The Trade-Off
v2.0 costs you in TWO ways:
- 💸 Money - Requires commercial license
- 🐌 Speed - 6-33x slower than v1.0 (but way smarter!)
| Feature | v1.0 (Free) | v2.0 (Premium) |
|---|---|---|
| Speed | Sub-3ms ⚡ | ~20-100ms 🐢 |
| Layers | 1 (semantic) | 5 (multi-dimensional) |
| Emotional Context | ❌ | ✅ |
| Temporal Awareness | ❌ | ✅ |
| Adaptive Retrieval | ❌ | ✅ |
| License | FREE | PAID |
Perfect For
✅ AI Agents - Need context, emotions, and timing
✅ Emotional AI - Chatbots with sentiment awareness
✅ Personal Knowledge - Systems that remember what matters AND when
✅ Intelligent Assistants - Context-aware retrieval
✅ Research Platforms - Multi-dimensional search
Get Your License
📧 Contact for pricing:
- Email: kent.stone@gmail.com
- Phone: +51 945 012 953
Pricing Tiers:
- Starter - Development & testing
- Pro - Production applications
- Enterprise - Unlimited usage
- Academic - Research discounts
Download & Run
- Contact us for a license key
- Download binary for your platform below
- Run with license:
TMC_LICENSE_KEY="your-key-here" ./tmc-server
API Endpoints
Basic Endpoints (v1.0 compatible):
POST /crystallize- Simple storagePOST /retrieve- Simple retrieval
Advanced Endpoints (v2.0 exclusive):
POST /v2/crystallize- Multi-layer storagePOST /v2/retrieve- Adaptive retrieval
Example Usage
Store memory with emotional context:
curl -X POST http://localhost:8000/v2/crystallize \
-H "Content-Type: application/json" \
-d '{
"text": "Feeling excited about the AI project\!",
"importance": 0.9,
"emotion_vector": [0.8, 0.1, 0.1],
"metadata": {"category": "work", "mood": "positive"}
}'Adaptive retrieval:
curl -X POST http://localhost:8000/v2/retrieve \
-H "Content-Type: application/json" \
-d '{
"query": "positive work experiences",
"k": 10,
"mode": "Adaptive"
}'Retrieval Modes
- Adaptive - AI decides the best strategy (recommended)
- SemanticOnly - Pure meaning-based search
- HashOnly - Exact keyword matching
- Hybrid - Custom weighted combination
- Temporal - Time-aware retrieval
- Emotional - Sentiment-based matching
Not Sure?
Start with TMC v1.0 (FREE) and upgrade when you need:
- Emotional intelligence
- Temporal context
- Adaptive retrieval
- Multi-layer analysis
Support
- 📧 Email: kent.stone@gmail.com
- 📞 Phone: +51 945 012 953
- 🐛 Issues: https://github.com/kentstone84/Ai_Memory/issues
© 2025 TMC. Proprietary. Commercial License Required.
TMC v1.0 - FREE & BLAZING FAST
TMC v1.0 - Lightning Fast Vector Search
255× faster than Pinecone. 100% FREE. No license required.
Performance (100k vectors)
- ⚡ Sub-3ms retrieval latency
- 📊 92% precision in similarity search
- 🚄 255× faster than Pinecone
- 💰 100% FREE - No license, no limits
- 🔓 Open Source - Use anywhere, anytime
What You Get
- Simple, blazing-fast vector similarity search
- Perfect for RAG, basic similarity search, and high-speed retrieval
- No licensing, no restrictions
- Production-ready performance
Download & Run
- Download binary for your platform below
- Make executable (Unix):
chmod +x tmc-server - Run:
./tmc-server - Start storing and retrieving at lightning speed!
API Endpoints
POST /crystallize- Store a memoryPOST /retrieve- Search memoriesGET /health- Health checkGET /stats- System statistics
Example Usage
Store a memory:
curl -X POST http://localhost:8000/crystallize \
-H "Content-Type: application/json" \
-d '{"text": "Your memory here", "importance": 0.8}'Retrieve similar memories:
curl -X POST http://localhost:8000/retrieve \
-H "Content-Type: application/json" \
-d '{"query": "search term", "k": 5}'Need More?
If you need emotional context, temporal awareness, and multi-dimensional search, check out TMC v2.0 (Premium).
Support
- 📧 Email: kent.stone@gmail.com
- 📞 Phone: +51 945 012 953
- 🐛 Issues: https://github.com/kentstone84/Ai_Memory/issues
© 2025 TMC. Open Source. No strings attached.