Benchmark Results

altimate-code publishes a score on every agentic data benchmark it has entered, across five database platforms, beating agents on more expensive models.

Benchmarked Across the Whole Data Stack

Platform coverage

Snowflake

DuckDB

PostgreSQL

MongoDB

SQLite

Tested and top-ranked across five database systems. Most data agents are built for a single platform. altimate-code is evaluated on all of them, with every score published.

Sonnet 4.6 · baseline

View full results →

dbt analytics-engineering tasks

Snowflake · DuckDB

Sonnet 4.6 + GPT-5.5-mini + Haiku 4.5

heterogeneous multi-database data tasks

PostgreSQL · MongoDB · SQLite · DuckDB

Source →

Note

Both benchmarks are independent, third-party evaluations — ADE-Bench from dbt Labs, DAB from UC Berkeley's EPIC Data Lab. altimate-code ships 100+ purpose-built, deterministic tools for data agents — SQL analysis, column-level lineage, dbt, FinOps, and warehouse connectivity — which is why it leads across platforms regardless of the backbone model.

Reproduce This on Your Warehouse. Same Agent, Your Data.

The same Altimate Code build that leads ADE-Bench and DAB runs on your own dbt project. Install it, or book a demo and we run it with you.

Get Started Book a Demo

Altimate Code: the skill browser in the terminal, a session recap and a dbt quality scan in VS Code