Self-hosted semantic layer for AI agents.
Best for Enabling AI agents to generate accurate, governed business reports from raw warehouse data.
Self-hosted semantic layer for AI agents.
Docs · CLI · Discord · Website.
What it does
- MCP server for AI agents to query semantic layers
- SQL-based metric definitions with caching and access control
- Multi-database support including Snowflake, BigQuery, and PostgreSQL
- Cube Store pre-aggregation cache for fast analytical queries
- Admin UI for browsing deployed models, views, and measures
Setup requirements
Requires 3 environment variables: ADMIN_TOKEN, CUBEJS_DB_TYPE, CUBEJS_DB_*. Available via NPX and Docker.
vs
A unified biomedical graph database that integrates 50+ primary data sources
Best for Retrieving high-resolution protein structures for specific gene targets.
*A unified biomedical graph database that integrates 50+ primary data sources — genes, proteins, compounds, diseases, pathways, and clinical data — into a single queryable graph with billions of cross-reference edges. Its native MCP server gives LLMs direct access to structured,…
What it does
- Integrates 50+ primary biomedical data sources including genes, proteins, and compounds
- Provides billions of cross-reference edges for complex biomedical queries
- Supports natural language querying for structured biomedical data
- Enables cross-database mapping (e.g., Ensembl to UniProt to PDB)
Biggest differences
CompareBonnardBioBTree
Best forEnabling AI agents to generate accurate, governed business reports from raw warehouse data.Retrieving high-resolution protein structures for specific gene targets.
StandoutMCP server for AI agents to query semantic layers.Integrates 50+ primary biomedical data sources including genes, proteins, and compounds.
SetupNPX or Docker, needs 3 env vars, stdio transport.Claude Desktop, http transport.
Transportstdiohttp
Community19 GitHub stars16 GitHub stars
Bottom line
Pick Bonnard if...Enabling AI agents to generate accurate, governed business reports from raw warehouse data. MCP server for AI agents to query semantic layers. NPX or Docker, needs 3 env vars, stdio transport.
Pick BioBTree if...Retrieving high-resolution protein structures for specific gene targets. Integrates 50+ primary biomedical data sources including genes, proteins, and compounds. Claude Desktop, http transport.
The real split here is workflow fit, not raw counts. Bonnard: Enabling AI agents to generate accurate, governed business reports from raw warehouse data. BioBTree: Retrieving high-resolution protein structures for specific gene targets. Public traction is fairly close (19 vs 16 stars).