Benchmark

Bird-SQL (dev)

BIRD (BIg Bench for LaRge-scale Database Grounded Text-to-SQLs) is a comprehensive text-to-SQL benchmark containing 12,751 question-SQL pairs across 95…

Modality
text
Categories
reasoning
Openness
unknown
Source
llm_stats
Reported scores
7

Reported scores

llm_stats

ModelOrganizationReported valueReportedEvidence
Gemini 2.0 FlashGoogle0.5692024-12-01evidence
Gemini 2.0 Flash-LiteGoogle0.5742025-02-05evidence
Gemma 3 12BGoogle0.4792025-03-12evidence
Gemma 3 1BGoogle0.0642025-03-12evidence
Gemma 3 27BGoogle0.5442025-03-12evidence
Gemma 3 4BGoogle0.3632025-03-12evidence
Nemotron 3 Super (120B A12B)NVIDIA0.4182026-03-11evidence

Scores are partitioned by the source that reported them and are never merged into a single cross-source ranking, because the sources measure different things and say so.