SQLMesh
Open-source data transformation with virtual environments and surgical backfills.
About
SQLMesh is an open-source SQL transformation framework developed by Tobiko Data as a dbt alternative. Its headline features are virtual data environments that give developers a full logical copy of production for testing, and automatic change categorization that uses SQLGlot-based parsing to decide whether an edit needs a backfill or just a metadata swap.
In September 2025 Fivetran acquired Tobiko Data to add AI-ready data transformation to its platform, and in March 2026 Fivetran contributed SQLMesh to the Linux Foundation under open governance. A month after the acquisition, Fivetran also announced an all-stock merger with dbt Labs, completed in June 2026, so one company now owns both major open-source SQL transformation frameworks, though both remained permissively licensed after the deals.
What it can do
- Virtual data environments for zero-cost dev testing
- Automatic breaking versus non-breaking change detection
- Interval-based incremental backfills
- Column-level lineage via SQLGlot parsing
- Engine support across Snowflake, BigQuery, Databricks, Postgres
Setup
pip install sqlmesh, define models in SQL or Python, and run sqlmesh plan to preview and apply changes.
Prerequisites
Python 3.9+; a data warehouse or DuckDB for local development.
Use cases
Works with
Trust score — 8/10
Scored 8/10 (verified tier, metrics-v1). Evidence: Verified via vendor site/docs/GitHub and reputable coverage, wave 90; Trending score 82/100 from researcher assessment.
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.