Databricks on track to pass Snowflake
Databricks at $4.8B ARR
Databricks is taking the lead because it is no longer just selling a warehouse, it is selling the whole data and AI workbench where engineers run Spark jobs, build notebooks, train models, and then add SQL and AI on top. That broader product surface is why it reached $4.8B run rate in September 2025 at 55% growth while Snowflake finished fiscal 2026 at $4.47B product revenue growing 30%, which makes a 2026 crossover a direct result of mix, not just momentum.
-
At similar scale, the growth gap is real and persistent. Databricks was at $2.4B ARR in June 2024 growing 60%, then $4.8B by September 2025 growing 55%. Snowflake went from $2.67B product revenue in fiscal 2025 to $3.46B in fiscal 2026, then $4.47B in fiscal 2027, with growth stepping down from 30% to 24% guidance and then 30% actual.
-
The product mix explains why Databricks can compound faster. About 60% of Databricks revenue still comes from its core data platform, but Databricks SQL and AI are each already at roughly $1B run rate. Snowflake remains much more concentrated in core warehouse consumption, with AI and app layers still small relative to the base business.
-
The workflow is different in practice. Snowflake is strongest when a team wants a clean SQL warehouse that analysts and BI tools consume. Databricks starts earlier in the pipeline, where data engineers ingest raw data, run Spark transformations, train models, and then serve analytics from the same environment, which creates more places to add spend per customer.
The next leg is a race to own both the data layer and the application layer. Databricks is pushing into databases, app deployment, and agent tooling, while Snowflake is broadening across engineering, AI, and applications. If current trajectories hold, Databricks becomes the largest independent data platform in 2026, and Snowflake has to win by deepening usage inside its installed base rather than by expanding faster than the market.