Open-source semantic layers and metrics stores in 2026

New this season

1. 1
   Mnemiq agenticfabriq

   Open-source text-to-SQL engine you tune and measure on your own database

   Building block Experimental You are building a natural-language query feature and want to measure accuracy on your schema.

   ★ 51 Apache-2.0 Python Active, pushed 14 days ago
2. 2
   QwenPaw-Data agentscope-ai

   Agentic enterprise data analytics: governed facts (DataBridge), reusable methodology (Skill-Hub), and controllable execution (Host)

   Framework or platform Experimental You research or prototype autonomous data analysis agents.

   ★ 184 Apache-2.0 Python Active, pushed 14 days ago

All open-source semantic layers and metrics stores, ranked

Complete systems you can adopt 1

1. 1
   Saiku spiculedata

   Open-source semantic layer: one cube for Excel (MDX/XMLA), dashboards, and AI agents (MCP). Mondrian + Apache Calcite.

   Complete system Usable Your organisation uses OLAP cubes and wants Excel pivot access plus a web front end.

   ★ 1.3k Apache-2.0 Java Active, pushed 16 days ago

Frameworks and platforms to build on 1

1. 1
   QwenPaw-Data agentscope-ai

   Agentic enterprise data analytics: governed facts (DataBridge), reusable methodology (Skill-Hub), and controllable execution (Host)

   Framework or platform Experimental You research or prototype autonomous data analysis agents.

   ★ 184 Apache-2.0 Python Active, pushed 14 days ago

Building blocks for your own system 6

1. 1
   MetricFlow dbt-labs

   MetricFlow allows you to define, build, and maintain metrics in code.

   Building block Production-ready Your data team uses dbt and wants one governed definition of each business metric.

   ★ 1.9k Apache-2.0 Python Active, pushed 20 days ago
2. 2
   SLayer MotleyAI

   An embeddable, expressive semantic layer for AI agents and humans

   Building block Experimental You are building AI analytics and want metrics defined once for agents to use.

   ★ 306 MIT Python Active, pushed 13 days ago
3. 3
   Data-tools Intugle

   The GenAI-powered toolkit for automated data intelligence.

   Building block Experimental You inherit scattered tables and want a first-pass semantic model generated automatically.

   ★ 152 Apache-2.0 Jupyter Notebook Active, pushed 2 months ago
4. 4
   Sidemantic sidequery

   The universal metrics layer. Compatible with 15+ formats: Cube, MetricFlow, LookML, Omni, BSL, LDM, Cortex, Malloy, OSI, SML, TML, Hex, Rill, Superset

   Building block Experimental You have metrics defined in one BI tool's format and want to query them consistently elsewhere.

   ★ 131 AGPL-3.0 Python Active, pushed 14 days ago
5. 5
   Zillion totalhack

   Make sense of it all. Semantic data modeling and analytics with a sprinkle of AI. https://totalhack.github.io/zillion/

   Building block Usable You want a lightweight semantic layer inside a Python application.

   ★ 208 MIT Python Slowing, pushed 4 months ago
6. 6
   Mnemiq agenticfabriq

   Open-source text-to-SQL engine you tune and measure on your own database

   Building block Experimental You are building a natural-language query feature and want to measure accuracy on your schema.

   ★ 51 Apache-2.0 Python Active, pushed 14 days ago

Developer tools 1

1. 1
   Dex exmergo

   Dex lets your agent build your analytics.

   Developer tool Experimental An analytics engineer wants an AI agent to do routine modelling work in their warehouse.

   ★ 105 Apache-2.0 Python Active, pushed 3 days ago

Lists, courses and other material 1

1. 1
   SML semanticdatalayer

   Open-source repository for Semantic Modeling Language (SML)

   List, course or other material Usable You want a vendor-neutral way to describe metrics and dimensions, or you use AtScale.

   ★ 177 Apache-2.0 Active, pushed 7 days ago

Open-source licenses

• Permissive: 9. Use it anywhere, including in closed products. You keep the copyright notice; Apache-2.0 also grants patent rights.
• Network copyleft: 1. Like GPL, and running a changed version as a web service counts as distributing it, so its users get the source too.

Languages

Python 7, Java 1 and Jupyter Notebook 1.