OpenLM Text2SQL - Spider EX: leaderboard
Execution accuracy on Spider's 10,181 cross-domain text-to-SQL pairs, measuring whether predicted SQL produces the same result as the gold query across 200+ databases.
Metric: Exact Execution (%). Source: openlm.ai. 10 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | MiniSeek | 91.2 | |
| 2 | DAIL-SQL + GPT-4 + Self-Consistency | 86.6 | |
| 3 | DAIL-SQL + GPT-4 | 86.2 | |
| 4 | DPG-SQL + GPT-4 + Self-Correction | 85.6 | |
| 5 | DIN-SQL + GPT-4 | 85.3 | |
| 6 | Hindsight Chain of Thought with GPT-4 | 83.9 | |
| 7 | C3 + ChatGPT + Zero-Shot | 82.3 | |
| 8 | Hindsight Chain of Thought with GPT-4 and Instructions | 80.8 | |
| 9 | RESDSQL-3B + NatSQ | 79.9 | |
| 10 | SeaD + PQL | 78.5 |
Interactive version: theaggregate.ai/benchmark?slug=openlm-text2sql-spider-ex · How It Works · Data refreshed daily, snapshot 2026-10-20.