NornicDB Hermes Memory CPU

NornicDB is a CPU-only, Neo4j-compatible graph, vector, and temporal database that serves as a long-term memory store for the Hermes agent.

  • AI
  • Databases
  • Network Services
  • Tools & Utilities
NornicDB is a CPU-only, Neo4j-compatible graph, vector, and temporal database that serves as a long-term memory store for the Hermes agent. It bundles BGE-M3 embeddings (no GPU required) and delivers sub-millisecond HNSW vector search, native graph traversal, and hybrid vector + BM25 retrieval in a single engine. It exposes Bolt/Cypher and Qdrant-compatible gRPC, so the agent can store, auto-link, and recall memories without a separate vector store. Managed embeddings, automatic relationships, MVCC temporal as-of reads, and policy-based memory decay let it build a persistent, self-organizing knowledge graph that forgets noise and promotes durable facts — a solid fit for Graph-RAG memory.

Platform availability

Available on 2 of 10 platforms

  • Unraid
  • TrueNAS (not listed)
  • Umbrel (not listed)
  • ZimaOS (not listed)
  • Proxmox (not listed)
  • Helm (not listed)
  • Mac
  • Windows (not listed)
  • Linux (not listed)
  • Any server (not listed)

Health score?A 0-100 score computed nightly from four factors: maintained (40%: recent commits and steady activity), popular (25%: GitHub stars, log scale), easy to install (20%: platforms, install notes, setup guides, Docker image) and light to run (15%: minimum RAM, ARM support). When the source code has been checked, a fifth factor, security check (20%: automated scan for committed secrets, unsafe desktop-app settings, known-vulnerable dependencies and project hygiene; any serious finding caps it at 20), is added. Factors without data are left out and the rest are rescaled.

78/100 Good

Maintained
100last commit 10 days ago
Popular
59930 GitHub stars
Easy to install
452 install platforms, install notes, Docker image
Light to run
95512 MB minimum RAM, runs on ARM
Security check
76no serious findings, 4 minor, CI, tests

Availability: platform listings confirmed today, links checked 34 days ago

GitHub stars
930
Open issues
16
Last commit
2026-11-11
Latest release
v1.4.3
License
MIT
Activity
declining

Checked 10 days ago - source: GitHub (orneryd/NornicDB)

Resources & Compatibility

Minimum requirements: 512 MB RAM. ARM (e.g. Raspberry Pi) is supported.

Checked 54 days ago - source: Docker Hub image tags and published docs

First-install notes

  • Install NornicDB as a service Download the NornicDB program file, make it runnable, then run `sudo nornicdb install`, `sudo systemctl start nornicdb`, and `sudo systemctl enable nornicdb` on your Linux machine. This keeps NornicDB running in the background and starts it automatically after a reboot, so you do not need to relaunch it every time you log in. source
  • Use the config and ports you need Start it with `nornicdb serve`, and if your setup uses custom settings, add `--config`, `--bolt-port 7687`, and `--http-port 7474` (Bolt is the database connection port, and HTTP is the web port). If those ports are blocked or the config file is missing, clients may not be able to connect to your database. source
  • Set OAuth login details first If you use the OAuth provider (a login/sign-in service), start it with the client ID and secret you plan to use, for example `./oauth-provider -client-id my-client -client-secret my-secret`, and pick a port like 8888 or 9999. This prevents login problems later, because missing or mismatched OAuth client details can stop users from signing in. source
  • Ports and data Expose ports 7474 (browser) and 7687 (Bolt/Cypher), and attach a persistent nornicdb-data volume so the database survives container restarts. source

Fetched 38 days ago - each note links to its source

Alternatives

No alternatives collected yet.

Common questions

Which Docker image should I use for a CPU-only NornicDB host, and which ports need to be exposed?

Use the CPU-only Docker image from the quick start for amd64 hosts, and expose 7474 for the web/admin UI and 7687 for the database connection.

Why does the Apple Silicon Docker image not provide GPU acceleration on macOS?

Docker on macOS does not expose Metal acceleration. The Apple Silicon image still runs, but GPU acceleration requires a native install from the releases page or a local build.

How do I keep NornicDB running automatically on Linux?

Download the program file, make it runnable, then run sudo nornicdb install, sudo systemctl start nornicdb, and sudo systemctl enable nornicdb. This runs it in the background and starts it after reboot.

What details must be set before using the OAuth login provider?

Start the OAuth provider with the intended client ID and client secret, and choose its port, for example 8888 or 9999. Missing or mismatched OAuth client details can prevent users from signing in.

Where can I find guidance for moving an existing Neo4j setup to NornicDB?

The documentation hub lists a Migrating from Neo4j section for that task.

Answers sourced from github.com, github.com, github.com, orneryd.github.io

Community

Collected 48 days ago - public community threads