OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Agent Memory & RAG

Papr Memory

Papr Memory is a self-hostable Python memory service for AI agents that stores documents and structured information and retrieves relevant memories through semantic search, vector embeddings and graph relationships.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Papr Memory — OpenViST project profile in Agent Memory & RAG.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use Papr Memory when an agent or application needs a separately operated store for persistent context, semantic search or graph-linked information, and you can manage its database and service dependencies.

Who it is for

  • Developers building AI agents that need persistent context across interactions.
  • Teams creating retrieval-augmented generation or knowledge-search systems.
  • Operators who want to self-host an agent memory service and its supporting databases.

Documented uses

  • Store text, documents, code snippets and structured data for later retrieval.
  • Search stored memories using natural-language queries and vector embeddings.
  • Represent connections between memories as graph relationships.
  • Provide a FastAPI-based memory service for client applications.
Before you choose

Requirements & limits

The repository identifies the project as open source under the AGPL-3.0 license and documents a self-hosted setup. The README lists managed infrastructure and automatic backups as cloud-only features.

  • The architecture uses MongoDB, Neo4j, Qdrant and Redis, so deployment entails operating several supporting services.
  • Local Qwen3 embeddings avoid external embedding API calls but have higher documented hardware requirements than the cloud-embedding option.
  • The README lists OpenAI credentials for LLM operations and a DeepInfra token for cloud embeddings.
  • The open-source edition and managed cloud offering have different documented features.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Set up Papr Memory with Docker Compose

Start the documented open-source services and locate the local API documentation.

Before you start

  • Python 3.8 or later
  • Git
  • Docker and Docker Compose
  • For local embeddings, the README recommends at least 8 GB RAM, 2 CPU cores, 2 GB swap and 10 GB disk.
  • An OpenAI API key is required for LLM operations; a DeepInfra token is needed only when using cloud embeddings.

Steps

  1. Clone the repository and change into its directory.

  2. Copy .env.example to .env.opensource and edit the configuration with the credentials needed for your chosen setup.

  3. Run the command below to start the open-source services.

    Documented command
    docker compose up -d

  4. Open the API documentation at http://localhost:5001/docs.

Expected outcome

The README says the open-source setup initializes its schemas and creates a default user and API key on first run; it lists the API documentation at the local address above.

Guide limitations

  • The guide documents service startup and API-documentation access, not an agent integration or a memory-write/search walkthrough.
  • Cloud embeddings require a DeepInfra token; local embeddings use the Qwen3 embedding model and have higher listed resource requirements.
  • Managed infrastructure, automatic backups and enterprise SSO are listed as cloud features, not open-source features.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
16 October 2026
Original website title
Papr-ai/memory-opensource · Official homepage title · retained in its source language
How we review sourcesSuggest a correction