Memory infrastructure for AI agents

Your agents lose everything when a session ends. Repogen stores memories with semantic search and cryptographic ownership, so your agents remember across sessions, frameworks, and deployments.

Get your API keyView source

What Repogen does

A simple API that lets your AI agents store and retrieve memories. When your agent learns something, store it. When your agent needs context, retrieve it.

Store memories

Save any text with optional metadata. We generate embeddings and index them for semantic search automatically.

Search by meaning

Query with natural language. Results are ranked by semantic similarity, not keyword matching.

Own your data

Every memory is signed with your DID key. You control who can read, write, or share your agent's memories.

Get started in three steps

Create an API key, install the SDK, and start storing memories. Takes less than five minutes.

1

Create your API key

Open the dashboard and generate a new API key. This key authenticates your agent with the Repogen network.

Open Dashboard
2

Install the SDK

Choose Python or TypeScript. Both SDKs provide the same functionality with native async support.

Python
pip install repogen
TypeScript
npm install @repogen/sdk
3

Store and retrieve memories

Initialize the client with your API key. Store memories when your agent learns something. Retrieve them when context is needed.

from repogen import RepogenClient

client = RepogenClient(api_key="your_api_key")

# Store a memory
await client.store(
    content="User prefers dark mode and minimal notifications",
    metadata={"type": "preference", "user_id": "123"}
)

# Retrieve relevant memories
memories = await client.search(
    query="what are the user's preferences",
    limit=5
)

# Use memories in your agent's context
for memory in memories:
    print(memory.content)

Built for autonomous agents

Whether you're building a personal assistant, a research agent, or a multi-agent system, Repogen handles the memory layer.

Personal assistants

Remember user preferences, past conversations, and accumulated context. Your assistant becomes more useful over time instead of starting fresh every session.

Research agents

Store findings, sources, and insights as your agent explores. Build a knowledge base that persists across research sessions.

Multi-agent systems

Share memories between agents with permission controls. Specialist agents can access shared knowledge while maintaining their own private memories.

LangChain integration

Drop-in memory backend for LangChain agents. Replace default memory with persistent, searchable storage in one line of code.

Start building with persistent memory

Repogen is open source. Run your own instance or use our hosted API. Get started by creating your API key in the dashboard.

Open DashboardView on GitHub