🧠 Remembra v0.8.0 — Production Ready

Semantic memory for AI agents that persists forever.

✨ What's New

🚀 Performance

• httpx Connection Reuse — Persistent HTTP clients eliminate 100-300ms TCP+TLS overhead per operation
• Optimized Entity Graph — Single endpoint call instead of N+1 queries

🔌 Multi-Provider Support

• Entity Extraction — Now supports OpenAI, Anthropic, and Ollama for entity extraction
• 6 Embedding Providers — OpenAI, Anthropic, Cohere, Voyage, Ollama, and custom endpoints

📊 Usage Tracking

• Warning Banners — Proactive alerts at 60%, 80%, and 95% usage
• X-Remembra-Usage-Percent — Header in all API responses

⚡ Quick Start DX

• One-liner install — curl -fsSL https://remembra.dev/install.sh | bash
• Zero-config Docker — docker-compose.quickstart.yml includes Qdrant + Ollama + Remembra

🧪 Testing

• 125 new tests — Total: 263 passed, 0 failures
• Full coverage — Embeddings, entities, conflicts, spaces, plugins

📦 Installation

# Python
pip install remembra

# TypeScript/Node
npm install remembra

# Docker
docker pull remembra/remembra:0.8.0

🔗 Links

• Landing: https://remembra.dev
• Docs: https://docs.remembra.dev
• Dashboard: https://cloud.remembra.dev
• Discord: https://discord.gg/Bzv3JshRa3

🙏 Launch Day: March 11, 2026

Give us a ⭐ if Remembra helps your AI remember!