ArcadeDB Python

Python bindings for ArcadeDB, the multi-model database (graph, document, key/value, vector and time series, queried with SQL or openCypher). The engine runs inside your Python process through JPype, with no server and no network hop. The wheel bundles its own Java 25 runtime, so you install no Java. Status: beta.

Install

pip install arcadedb   # or: uv add arcadedb

This package was published as arcadedb-embedded until 26.10.x; install arcadedb now.

• Python 3.11 to 3.14. CI builds and tests all four versions on every platform below for main, every night and each release. A pull request runs a smaller selection.
• Linux x86_64 and ARM64, macOS on Apple Silicon, Windows x86_64. One wheel per platform and Python version.
• Import name: arcadedb. Optional extras: arrow, pandas, vector.

Quick start

import arcadedb

with arcadedb.create_database("./mydb") as db:
    db.command("sql", "CREATE DOCUMENT TYPE Person")
    db.command("sql", "CREATE PROPERTY Person.name STRING")
    db.command("sql", "CREATE PROPERTY Person.age INTEGER")

    # Writes need a transaction
    with db.transaction():
        db.command("sql", "INSERT INTO Person SET name = ?, age = ?", "Alice", 30)

    for record in db.query("sql", "SELECT FROM Person WHERE age > 25"):
        print(f"Name: {record.get('name')}")

This prints Name: Alice (the engine also logs to standard error). create_database raises if the path already holds a database, so run it in an empty directory. Everything else goes through the same two calls, db.command(...) and db.query(...), in SQL or openCypher.

What is included

• Embedded mode: the database is a directory, opened by one process at a time.
• Graph, document, key/value, vector (JVector HNSW) and time series models; ACID transactions.
• Import from CSV, XML and ArcadeDB JSONL; export to JSONL.
• Columnar results with to_columns(), to_arrow() and to_dataframe().
• An optional in-process HTTP server (create_server()), off until you call it, for other processes that need the same data. From 26.11.1 the wheels also carry the Studio web UI, which the server then serves. Studio used to bundle ApexCharts, which is no longer open source (ArcadeData/arcadedb#10579), and ArcadeData/arcadedb#10627 replaced it with Apache ECharts. The earlier 26.10.x wheels on PyPI have no Studio.

For several processes, other languages, HA or replication, run the regular ArcadeDB server and connect with arcadedb-driver (HTTP client, developed in ArcadeData/arcadedb-drivers).

Packages on PyPI

Package	What it is	Status
arcadedb	These bindings: the engine on a bundled JVM. The package this README describes	First release from this repository: 26.11.1
arcadedb-native	The same API with the engine compiled ahead of time by GraalVM Community Edition: no JVM, no JPype. Python 3.11 or later. It imports as arcadedb, like the JVM wheel, so install only one of the two	Not in this repository yet (porting is tracked in #19). Until then a placeholder on PyPI
Formerly arcadedb-embedded and arcadedb-embedded-native	The names of these packages until 26.10.x. They import as arcadedb_embedded and arcadedb_native	Deprecated: install arcadedb

Releases from this repository start with 26.11.1. The 26.10.x wheels were built and published from humemai/arcadedb-embedded-python.

Documentation

• Here: docs/ is an Antora component (AsciiDoc) with Installation, Which package for which mode, Quickstart, Databases and queries, and Concurrency, processes and crashes. It covers only what is specific to the Python bindings.
• The component is not published on docs.arcadedb.com yet; adding it as a content source is open in ArcadeData/arcadedb-docs#505.
• The guide, API reference and example walkthroughs for the published package are still at docs.humem.ai/arcadedb until those pages are ported here (#19).
• ArcadeDB itself (SQL, openCypher, schema, server): docs.arcadedb.com. It is not repeated here.

Examples

examples/ holds 27 runnable scripts, from a first document store to vector search, time series, graph algorithms and server mode. CI runs them (test-python-examples.yml).

Build and test from source

./scripts/build.sh   # a wheel for your platform in dist/

Linux builds run in Docker. macOS and Windows need a JDK 25 or later. The jars come from the arcadedata/arcadedb Docker image: latest by default, or the tag in ARCADEDB_TAG (for example ARCADEDB_TAG=26.10.2 ./scripts/build.sh). The package version is read from the engine jar in that image, so a 26.11.1-SNAPSHOT image builds 26.11.1.devN.

Tests run against the built wheel, not the source tree: install the wheel with its test extra into a virtual environment, then run pytest. CLAUDE.md has the exact commands, the build options and the lint setup.

Contributing

Open an issue or a pull request. Run pre-commit run --all-files first, and add a regression test with every fix. The rest of the bindings work from humemai/arcadedb-embedded-python is ported here one pull request at a time; #19 lists what is done and what is next.

Where this code comes from

It moved here from bindings/python in ArcadeData/arcadedb at ee04bd202abce0183c6263a3c4d8a61a4d526960, with its git history. Issue numbers in older commit messages refer to that repository. humemai/arcadedb-embedded-python is the fork where the bindings kept developing (runtime fixes, JRE trim, native build, benchmarks). Its benchmark tooling stays there.

Citation

If you use these bindings in published work, cite them with the metadata in CITATION.cff; GitHub shows it as "Cite this repository". There is no DOI for this repository yet.

License

Apache License 2.0, like ArcadeDB. See LICENSE.