Memos

A flexible and extensible memory management system that supports multiple LLMs, vector stores, and embedding providers.

Features

• 🧠 Multiple LLM Support (OpenAI, Anthropic, Ollama)
• 🔤 Embedding Providers (OpenAI, Ollama)
• 💾 Vector Store Integrations (Qdrant, ChromaDB)
• ⚙️ Flexible Configuration & Extensible Architecture

Quick Start

from aios.memos.core import Memos, MemoryConfig, VectorStoreConfig, LlmConfig, EmbedderConfig

# Configure
config = MemoryConfig(
    vector_store=VectorStoreConfig(provider="qdrant"),
    llm=LlmConfig(provider="openai"),
    embedder=EmbedderConfig(provider="openai")
)

# Initialize
memos = Memos(config)

# Add memory
memory = memos.add("The sky was beautiful today", {"tag": "observation"})

# Search
results = memos.search("What about the sky?")

Installation

pip install aios-memos

License

MIT License