Skip to content

Remembra

Persistent memory for AI applications. Self-host in 5 minutes.

What is Remembra?

Remembra is a universal memory layer for LLMs. It solves the fundamental problem that every AI forgets everything between sessions.

from remembra import Memory

memory = Memory(user_id="user_123")

# Store memories
memory.store("User prefers dark mode and works at Acme Corp")

# Recall with context
context = memory.recall("What are user's preferences?")
# Returns: "User prefers dark mode. Works at Acme Corp."

Why Remembra?

The Problem

Every AI app needs memory. Developers hack together solutions using vector databases, embeddings, and custom retrieval logic. It's complex, fragmented, and everyone rebuilds the same thing.

Current Solutions

  • Mem0: $24M raised, but self-hosting docs are trash, pricing jumps from $19 to $249
  • Zep: Academic, complex to deploy
  • Letta: Not production-ready
  • LangChain Memory: Too basic, no persistence

Our Approach

  • Self-host in 5 minutes: One Docker command, everything bundled
  • Fair pricing: $0 → $29 → $99 (not $19 → $249)
  • Open source core: MIT license, own your data
  • Actually works: Built because we need it ourselves

Core Features

  • :material-brain:{ .lg .middle } Smart Extraction

    LLM-powered fact extraction transforms messy conversations into clean, searchable memories.

  • :material-account-group:{ .lg .middle } Entity Resolution

    Knows that "Adam", "Mr. Smith", and "my husband" are the same person.

  • :material-clock-time-four:{ .lg .middle } Temporal Memory

    TTL support, memory decay, and historical queries with as_of.

  • :material-magnify:{ .lg .middle } Hybrid Search

    Semantic + keyword search with CrossEncoder reranking for accurate recall.

  • :material-graph:{ .lg .middle } Entity Graph

    Traverse relationships to find related memories across your knowledge graph.

  • :material-docker:{ .lg .middle } Self-Host First

    One Docker command. All dependencies bundled. Your data stays yours.

Quick Start

docker run -d -p 8787:8787 remembra/remembra

pip install remembra

git clone https://github.com/remembra/remembra
cd remembra
docker-compose up -d

Then use the Python SDK:

from remembra import Memory

memory = Memory(
    base_url="http://localhost:8787",
    user_id="user_123"
)

# Store a memory
memory.store("User's name is John. He's a software engineer at Google.")

# Recall memories
context = memory.recall("Who is the user?")
print(context)  # "John is a software engineer at Google."

Get Started :material-arrow-right: View on GitHub :material-github:

Architecture

┌─────────────────────────────────────────────────────────────┐
│                     Your Application                         │
├─────────────────────────────────────────────────────────────┤
│                   Remembra SDK / API                         │
├──────────────┬──────────────┬───────────────┬───────────────┤
│  Extraction  │   Entities   │    Retrieval  │   Temporal    │
│  (LLM-based) │ (Resolution) │(Hybrid Search)│  (TTL/Decay)  │
├──────────────┴──────────────┴───────────────┴───────────────┤
│                      Storage Layer                           │
│         Qdrant (vectors) + SQLite (metadata/graph)          │
└─────────────────────────────────────────────────────────────┘

License

Remembra is open source under the MIT License.

Built with :heart: by DolphyTech