cloudparts

MindsDBDormant

A federated SQL engine that queries over 200 databases, warehouses, apps and files in place, and adds knowledge bases: tables you INSERT documents into and SELECT from by meaning, for retrieval-augmented generation.

What it is
Query engines
Used for
Retrieval-augmented generation (RAG)
Type
Open source
Made by
MindsDB, Inc.
License
Elastic License 2.0 (source-available, not OSI open source); the integration handlers are MIT
Status
Dormant. MindsDB now builds MindsHub, an AI agent workspace, and leaves the query engine to the community; its last release is 26.1.0 (April 23, 2026).
Facts checked
23 October 2026
Website Source code

What it is

MindsDB, now called the MindsDB Query Engine, is a Python server that puts one SQL dialect over many data sources. CREATE DATABASE attaches a source through an integration handler: PostgreSQL, MySQL, MongoDB, Snowflake, BigQuery, S3, and apps such as Slack, email or calendars. A query planner translates each query, pushes work down to the sources, joins across them and streams the results back, so nothing is copied; you can also send a source its own native query. The server speaks the MySQL and PostgreSQL wire protocols and has an HTTP API with a built-in SQL editor, so mysql, psql, DBeaver, SQLAlchemy or a BI tool can connect to it like an ordinary database.

Knowledge bases are its semantic layer. A knowledge base combines an embedding model, an optional reranking model and a vector store such as pgvector; INSERT INTO a knowledge base chunks, embeds and indexes rows, and SELECT ... WHERE content = '...' returns the closest chunks with a relevance score, filtered by metadata columns like any table. Hybrid search runs vector similarity and BM25 keyword matching together, for queries that mix plain language with exact codes. Views save cross-source queries, jobs rerun SQL on a schedule to keep knowledge bases fresh, and agents answer natural-language questions over the attached data using models from OpenAI, Anthropic, Google, Bedrock, Ollama and others.

In 2026 the company renamed its product MindsHub, an agent workspace with a model router and the open-source Anton agent, and its old GitHub repository became MindsHub’s. The engine moved to its own repository, mindsdb/engine, as a standalone open-source project that the company says is maintained by the community. Its last release, 26.1.0, came out on April 23, 2026, and later commits are mostly CI housekeeping. mindsdb.com now redirects to mindshub.ai.

Why teams pick it

  • One SQL interface over databases, warehouses, SaaS apps and files, queried live without an ETL pipeline
  • Retrieval in plain SQL: semantic, keyword and hybrid search with metadata filters, next to joins over live data
  • Speaks the MySQL and PostgreSQL protocols, so existing clients, ORMs and BI tools connect unchanged
  • Runs from one Docker image or pip install mindsdb

Watch out for

  • No longer the company’s product: development has largely stopped since April 2026, and support is community-only
  • Elastic License 2.0, not open source: you can’t offer it to others as a managed service
  • Federated queries are only as fast as the sources, and joins across sources run in the MindsDB server, so large ones move a lot of data through it
  • The product changed direction several times, from machine learning in the database to federated queries to agents, so older tutorials often don’t match the current engine

Specs

Compare with alternatives →
Data model
Federated tables over 200+ sources, plus knowledge bases searched by meaning
Query language
SQL in its own dialect, or a source’s native query
Client protocol
MySQL and PostgreSQL wire protocols, and an HTTP API
Written in
Python
Ways to run it
Self-hosted server; community-maintained since April 2026

Alternatives

What all query engines have in common →Compare side by side →
TrinoOpen sourceA distributed SQL engine that queries data where it lives, from Iceberg tables on S3 to a live PostgreSQL database, in one query.Spice.ai OSSOpen sourceA single-binary SQL engine in Rust, built on DataFusion and Arrow, that federates queries across databases, warehouses and lakes, keeps hot datasets in a local accelerated copy, and adds search and LLM APIs on top.ShillelaghOpen sourceA Python library and CLI that lets you query APIs, files and in-memory objects with SQL, by exposing each one as an SQLite virtual table; it is how Apache Superset reads Google Sheets.Presto (PrestoDB)Open sourceFacebook’s distributed SQL engine, now governed by the Linux Foundation, moving its workers from Java to C++ on Velox. Trino is its 2019 fork.Amazon AthenaAWSAWS’s serverless SQL engine for data in S3: no cluster to run, and on-demand queries cost $5 per TB scanned.Apache Arrow AceroExperimentalOpen sourceThe streaming execution engine inside Arrow’s C++ library: you hand it an exact plan of nodes (scan, filter, project, aggregate, join, sort) and it streams Arrow record batches through them, with no SQL parser or optimizer of its own.Apache CalciteOpen sourceA Java framework with everything a database needs to understand a query (SQL parser, validator, relational algebra and a rule-based, cost-based optimizer) but no storage of its own, which Hive, Flink, Druid, Dremio, Beam and many other systems embed as their SQL front end and planner.Apache DataFusionOpen sourceA query engine library in Rust, built on Arrow, that other projects embed and extend to build their own databases and data tools.+1 Apache DataFusion CometApache DrillOpen sourceA distributed SQL engine, inspired by Google’s Dremel, that queries JSON, Parquet, CSV files and NoSQL stores in place without defining a schema first.Apache HadoopOpen sourceThe project that made big-data clusters normal: a distributed file system (HDFS), a cluster scheduler (YARN) and the MapReduce engine.Apache HiveOpen sourceThe first SQL engine for Hadoop, and the source of the Hive Metastore, the table catalog that Trino, Spark, Impala and Flink still read.Apache ImpalaOpen sourceAn MPP SQL engine in C++ for interactive queries over Hive tables on HDFS, Kudu, Iceberg and object storage; built and mainly sold by Cloudera.Apache PigDormantOpen sourceA step-by-step scripting language, Pig Latin, that compiles to MapReduce, Tez or Spark jobs on Hadoop. Still released, but the project calls itself dormant.Apache SparkOpen sourceThe general-purpose distributed engine for large batch jobs, streaming and machine learning, in Python, SQL, Scala, Java or R.+1 Apache GlutenApache ZeppelinOpen sourceA self-hosted web notebook where each paragraph runs through an interpreter: Spark, Flink, JDBC, Python and more. Maintained, but behind Jupyter and Databricks in reach.chDBOpen sourceThe ClickHouse engine packaged as a library: run ClickHouse SQL inside a Python, Go, Rust or Node.js program, with no server.Cloudflare Basin SQLRenamedCommercialCloudflare’s serverless, distributed SQL engine for Apache Iceberg tables in Basin Catalog on R2, called R2 SQL until it became generally available on 1 October 2026, billed by the data it scans.DuckDBOpen sourceAn in-process analytical SQL database: a library with no server that queries Parquet, CSV, Iceberg and S3 directly.+1 DuckFlightGizmoSQLOpen sourceA database server that wraps DuckDB (or SQLite) and serves it to many clients over Arrow Flight SQL, with TLS and authentication.GlareDBDormantOpen sourceA small, embeddable SQL engine in Rust for querying Parquet and CSV files locally, over HTTP or in S3 and GCS, from a CLI, Python or the browser. It began as a DataFusion-based federated query service and was rewritten in 2025 on its own engine, and it has had no release since June 2025.HotdataCommercialA managed SQL engine for AI agents, built in Rust on Apache DataFusion and Arrow: each agent gets short-lived “instant databases” to load data into, then runs read-only SQL, vector, full-text and geospatial queries over an HTTP API or CLI.IbisOpen sourceA Python dataframe API that compiles to SQL, so the same code runs on DuckDB on a laptop and on BigQuery, Snowflake, PostgreSQL or more than 20 other engines.PolarsOpen sourceA DataFrame library with a Rust query engine underneath: you write Python (or Rust, SQL, Node.js, R) and it plans, optimises and runs the query on every core, streaming data bigger than memory.RayforceOpen sourceA small embeddable engine in pure C that runs columnar analytics, graph traversals and recursive Datalog queries in one optimised plan.SailOpen sourceA Spark Connect server written in Rust on Apache DataFusion: your PySpark code connects to it as if it were Spark, and it runs the queries without a JVM, on one machine or on Kubernetes.VeloxOpen sourceMeta’s open-source C++ execution engine library: it takes an already-optimized query plan and runs it with vectorized operators, and it is the engine inside Presto’s C++ workers (Prestissimo) and Apache Gluten’s native Spark.

How it fits with other technologies

Join two sources, then search support tickets by meaning · sql

CREATE DATABASE my_pg WITH ENGINE = 'postgres',
PARAMETERS = {"host": "localhost", "port": 5432, "user": "user", "password": "pass", "database": "mydb"};

-- one query across MongoDB and PostgreSQL, nothing copied
SELECT p.name, COUNT(t.ticket_id) AS open_tickets
FROM my_mongo.support_tickets AS t
JOIN my_pg.products AS p ON t.product_id = p.id
WHERE t.status = 'open'
GROUP BY p.name;

-- index ticket text in a knowledge base, then retrieve by meaning
INSERT INTO support_kb
  SELECT ticket_id, subject, body, product_name, priority, created_at
  FROM my_mongo.support_tickets;

SELECT chunk_content, product_name, relevance
FROM support_kb
WHERE content = 'cannot connect after the latest update'
  AND priority <= 2
LIMIT 10;

See it in a real stack

No architecture uses MindsDB yet. Drop it into the playground and wire it up yourself.Open playground →
Where it came from2017–2026

Jorge Torres and Adam Carrigan founded MindsDB in Berkeley, California in 2017. Its first open-source release, in 2018, brought basic machine-learning functions into traditional databases, and the name described that architecture. The company went through Y Combinator’s Winter 2020 batch and raised more than $55 million from investors including Benchmark, Mayfield, Y Combinator and NVIDIA.

As large language models arrived, the focus moved from predictions to data access for AI. Between 2023 and 2024 MindsDB added knowledge bases, semantic search and federated queries across documents and other unstructured sources, and it described itself as an engine for AI agents to query enterprise data. In 2025 it launched a separate agents product, and in 2026 it renamed its product MindsHub, keeping MindsDB as the company name. The engine’s last releases, 26.0 in February and 26.1 on April 23, 2026, came from the original repository; the next day its code got a repository of its own, mindsdb/engine, and was left to its community.

Why it’s built this way

Speak SQL and existing wire protocols
Every analyst and BI tool already speaks SQL and connects to MySQL or PostgreSQL. Presenting models, apps and documents as tables let them use AI features without new client libraries.
Query sources in place
Copying data into one store before using it means pipelines to build and data that goes stale. Pushing queries down to each source keeps answers current, at the cost of depending on each source’s speed.
Knowledge bases as tables
Retrieval for LLMs usually needs a separate vector database and indexing code. Making the embedding, storage and search a table you INSERT into and SELECT from keeps RAG inside the same SQL as everything else.

Timeline

  1. 2017

    Founded in Berkeley by Jorge Torres and Adam Carrigan

  2. 2018

    First open-source release: machine learning inside the database

  3. 2020

    Y Combinator Winter 2020 batch

  4. 2023

    Knowledge bases, semantic search and federated queries over unstructured data (2023–2024)

  5. 2025

    MindsHub Agents launched as a separate product

  6. 2026

    Product renamed MindsHub; the engine moves to its own repository, last release 26.1.0 (April 23)

Sources7 · facts checked 23 October 2026

Improve this page on GitHub