Benchmark RadarAI BENCHMARK PROFILE
UniQL
UniQL is a human-verified benchmark for cross-dialect text-to-SQL, aligning 1,534 questions with executable SQL across 16 dialects, with dialect-specific evaluation tracks.
- Released
- 2026-06-06
- Readiness
- Runnable
- Primary field
- General AI
Why it matters
Most text-to-SQL benchmarks are SQLite-only, but real systems use varied dialects; UniQL enables controlled evaluation of dialect generalization, highlighting transfer gaps.
Motivation
Existing text-to-SQL benchmarks are largely centered on SQLite, making it difficult to evaluate whether models can generalize across heterogeneous SQL dialects.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.