Omnigraph

Omnigraph

Open source graph database for your AI agents

Open-source graph database that gives AI agents one typed, branching knowledge graph with Git-like review and merge.

Omnigraph homepage

Omnigraph homepage

1 / 3

Overview

What is Omnigraph?

Open source graph database for your AI agents

Omnigraph is an open-source, object-storage native graph database built for AI agents. It allows AI agents to safely collaborate on shared knowledge: give agents one typed graph with Git-like branching, review, and merge for every write, so changes can be validated before becoming shared context.

Run it on your own S3-compatible storage, with semantic, keyword, and hybrid search plus graph traversal in one query. No hosted memory API or proprietary storage required. The engine, CLI and server are MIT licensed, and Omnigraph Studio offers a GPU-rendered desktop IDE for exploring the graph. A managed Cloud is in private preview.

Platforms and languages

Omnigraph Availability

Platforms

WebMacosWindowsLinuxApi

Languages

English

Capabilities

Omnigraph Key Features

Typed schema

Every agent shares one domain model, and every write is validated against the schema.

Git-like branching

Agents write on their own branches, then merge into main after review, with full history and time travel.

Hybrid search

Semantic, keyword and graph traversal queries run together in a single query.

Declarative policies

Cedar policies enforced inside the engine give fine-grained control over agent and human permissions.

Object-storage native

Data lives as open Lance files in your own S3-compatible storage, Azure Blob or local disk, with no lock-in.

Best for

Who uses Omnigraph?

Developers and AI engineers building agent systems

Give agents a shared, reviewable memory and context graph

Plans and access

Omnigraph Pricing

Free

Free and open source (MIT); Cloud in private preview

No free trial listed

Common questions

Omnigraph FAQs

Is Omnigraph open source?

Yes. The engine, CLI and server are MIT licensed, and graphs are stored as open Lance files.

How is it different from a vector database?

Vector databases find similar text; a graph database tracks relationships between people, decisions and events. Agents usually need both.

Where does my data live?

In your own storage — S3-compatible storage, Azure Blob or a local disk. The self-hosted engine itself is free.

Does it support MCP?

Yes. Omnigraph ships an MCP server alongside TypeScript and Python SDKs, so agents can query the graph directly.

Reviews

See what the community thinks and share your experience.

—

Based on 0 ratings

Rating distribution

0
0
0
0
0

Leave a review

Sign in to rate

Community reviews

No reviews

No written reviews yet.

Explore by category

AI Database Solutions

View all AI Database Solutions websites

Explore by category

AI Development Infrastructure

View all AI Development Infrastructure websites
Plurai preview
Plurai

Simulation and evaluation platform that tests AI agents in sandboxed digital twins, finds failures before users do, and closes the fix loop.

AI Development Infrastructure