At a glance
| Supported data platforms | 23 connectors: Snowflake, Databricks, BigQuery, Redshift, PostgreSQL, MySQL, MariaDB, SQL Server, Oracle, Fabric, Synapse, Teradata, SAP HANA, Db2, Netezza, Vertica, Trino and more |
| Anomaly detection | ML-driven anomaly detection monitors |
| Freshness and volume monitoring | Not documented in the pages checked |
| Schema checks | Not documented in the pages checked |
| Validation rules | Data diffs and data test monitors (YAML, run_in_ci flag) |
| Lineage and root cause | Downstream impact of changed dbt models, including BI assets |
| Alerting | Monitor verdicts and pull request comments |
| Deployment | SaaS (US, EU) or single-tenant VPC |
| Licensing | Commercial, quote only; migrations priced per object |
Datafold is strongest when you need to prove two datasets match: before and after a dbt change, or between a legacy source and a new warehouse during migration. Check which app you would be on, because its documentation says CI/CD testing is available only in Datafold Classic, and note that the free data-diff tool many teams used is archived.
Overview
Datafold compares data. A data diff checks tables, query results or files row by row and column by column, within one database or across two different ones, and reports where rows and values differ. Data diff monitors run these comparisons on a schedule and return an OK or Alert verdict, data test monitors can run in CI, and an anomaly detection feature monitors quality metrics. The documentation lists 23 connectors, from Snowflake, Databricks and BigQuery to Teradata, Netezza, Db2 and Oracle.
In CI, Datafold reads the production and pull request dbt manifest.json files, diffs the changed models and their downstream assets and posts the result as a pull request comment, with an optional AI code review and summary; non-dbt pipelines use an API or SQL parsing. The documentation says this CI/CD testing is only in Datafold Classic. The company also sells data platform migrations delivered by its Migration Agent, which combines AI code translation with diff-based validation, and exposes diffs to AI coding agents through an MCP server.
Key features
- Data diff within or across databases, at row and column level, including files in cloud storage
- Scheduled data diff monitors with OK or Alert verdicts, and data test monitors
- CI/CD testing for dbt and other pipelines with pull request comments (Datafold Classic only)
- AI code reviews and AI summaries of data impact on GitHub, GitLab and Bitbucket pull requests
- Migration Agent for data platform migrations with diff-based parity checks
- MCP server and API so coding agents and tools can run diffs
- VPC deployment in AWS, GCP or Azure
Integrations and ecosystem
Pros and cons
Pros
- Cross-database diffing is useful for migration validation
- Wide connector list including legacy warehouses such as Teradata and Netezza
- Shows data impact of code changes before merge (Classic app)
- VPC deployment option
Cons
- No public pricing or trial
- CI/CD testing documented as available only in Datafold Classic, not the current app
- Open source data-diff is archived, so there is no free self-hosted option from Datafold
- Not a full observability suite: freshness and volume monitoring are not its focus
Pricing
Datafold publishes no prices: its /pricing URL redirects to a contact page where pricing is discussed in a demo. For data platform migrations, Datafold says pricing is based on the number of objects, with price, timeline and quality guaranteed in the contract. No free tier or self-service trial is advertised, and the former free open source data-diff tool was archived in May 2024.
Pricing checked on the vendor's official page on 7 November 2026. Prices change; confirm before buying.
Who it is best for
Teams validating a warehouse migration or reviewing dbt changes who need precise row-level comparisons between datasets.
Typical use cases
- Prove parity between Teradata or Oracle tables and their Snowflake or Databricks copies during migration
- Show the row-level effect of a dbt model change in the pull request before merge
- Run a scheduled diff between a source system and a replicated table
- Let an AI coding agent verify its SQL changes against production data via MCP
Who should look elsewhere
- Teams that only need assertions inside dbt: see dbt tests.
- Teams looking for broad warehouse observability with freshness and volume alerts: see Monte Carlo.
Verdict
Datafold is a specialist in data diffing, valuable for migrations and for reviewing transformation changes with row-level evidence. Buyers should confirm whether CI/CD testing is available in the app version they would get, and budget for quote-based pricing now that the free data-diff tool is archived.
Frequently asked questions
Is data-diff still open source?
The datafold/data-diff repository remains readable under the MIT licence, but Datafold archived it on 17 May 2024 and no longer supports or develops it. Data diffing continues in the paid Datafold product.
How is Datafold different from dbt tests?
dbt tests assert rules (not null, unique, accepted values) on one version of a model. Datafold compares two versions of the data, such as production and a pull request build, or a source and a migrated target, and shows which rows and columns differ.
How much does Datafold cost?
Datafold publishes no prices; its pricing URL redirects to a contact form. Migrations are priced on the number of objects.
Does Datafold run in CI?
Yes, with GitHub, GitLab or CircleCI and dbt manifests, but its documentation says CI/CD testing is available only in Datafold Classic, not in the current Datafold app.
Sources
- Datafold home page
- Datafold database connections (documentation)
- How Datafold in CI works (documentation)
- Archived data-diff repository
- Sunsetting open source data-diff (Datafold blog)
- Datafold documentation home
Checked November 2026.
Listed in: Data Quality & Observability.
How we research reviews and what star ratings mean: our editorial method.