| A Preview of XiYan-SQL: A Multi-Generator Ensemble Framework for Text-to-SQL | ✓ Link | 75.63 | 73.34 | | XiYan-SQL | 2024-11-13 |
| []() | | 74.12 | 74.32 | | DSAIR + GPT-4o | |
| CHASE-SQL: Multi-Path Reasoning and Preference Optimized Candidate Selection in Text-to-SQL | | 74.06 | 73.14 | | CHASE-SQL + Gemini | 2024-10-02 |
| []() | | 73.17 | 72.43 | | ExSL + granite-34b-code | |
| []() | | 72.28 | 69.3 | | OpenSearch-SQL+ v2 + GPT-4o | |
| The Death of Schema Linking? Text-to-SQL in the Age of Well-Reasoned Language Models | | 71.83 | 67.21 | | Distillery + GPT-4o | 2024-08-14 |
| []() | | 70.26 | 72.16 | | Insights AI | |
| []() | | 70.21 | 68.12 | | PURPLE + RED + GPT-4o | |
| []() | | 69.40 | 68.91 | | MCTS-SQL | |
| []() | | 69.03 | 66.95 | | RECAP + Gemini | |
| []() | | 68.87 | 65.45 | | ByteBrain | |
| []() | | 67.86 | 65.38 | | ExSL + granite-20b-code | |
| CHESS: Contextual Harnessing for Efficient SQL Synthesis | ✓ Link | 66.69 | 65 | | CHESS | 2024-05-27 |
| []() | | 66.21 | 67.99 | | Arcwise + GPT-4o | |
| []() | | 65.45 | 63.36 | | MCS-SQL + GPT-4 | |
| []() | | 65.23 | 64.73 | | SCL-SQL | |
| []() | | 64.95 | 61.34 | | OpenSearch-SQL v1 + GPT-4 | |
| []() | | 64.84 | 60.5 | | PB-SQL v1 | |
| []() | | 64.51 | 62.97 | | PURPLE + GPT-4o | |
| []() | | 64.00 | 66.82 | | MSL-SQL + DeepSeek-V2.5 | |
| []() | | 63.39 | 55.48 | | SENSE-13B | |
| []() | | 63.39 | 55.48 | | SENSE | |
| []() | | 63.22 | 62.58 | | GRA-SQL | |
| []() | | 62.66 | 58.5 | | SuperSQL | |
| []() | | 60.71 | 59.71 | | Dubo-SQL, v1 | |
| []() | | 60.37 | 58.47 | | SFT CodeS-15B | |
| MAC-SQL: A Multi-Agent Collaborative Framework for Text-to-SQL | ✓ Link | 59.59 | 57.56 | | MAC-SQL + GPT-4 | 2023-12-18 |
| []() | | 59.25 | 57.17 | | SFT CodeS-7B | |
| Text-to-SQL Empowered by Large Language Models: A Benchmark Evaluation | ✓ Link | 57.41 | 54.76 | | DAIL-SQL + GPT-4 | 2023-08-29 |
| DIN-SQL: Decomposed In-Context Learning of Text-to-SQL with Self-Correction | ✓ Link | 55.90 | 50.72 | | DIN-SQL + GPT-4 | 2023-04-21 |
| Can LLMs Effectively Leverage Graph Structural Information through Prompts, and Why? | ✓ Link | 54.89 | 46.35 | | GPT-4 (Baseline) | 2023-09-28 |
| Can LLMs Effectively Leverage Graph Structural Information through Prompts, and Why? | ✓ Link | 49.02 | 42.70 | | Claude-2 (Baseline) | 2023-09-28 |
| []() | | 47.74 | 37.68 | | Open SQL-7B | |
| Can LLM Already Serve as A Database Interface? A BIg Bench for Large-Scale Database Grounded Text-to-SQLs | ✓ Link | 40.08 | 36.64 | | CoT + ChatGPT | 2023-05-04 |
| Can LLM Already Serve as A Database Interface? A BIg Bench for Large-Scale Database Grounded Text-to-SQLs | ✓ Link | 39.30 | 37.22 | | ChatGPT (Baseline) | 2023-05-04 |
| Can LLM Already Serve as A Database Interface? A BIg Bench for Large-Scale Database Grounded Text-to-SQLs | ✓ Link | 36.47 | 34.35 | | Codex (Baseline) | 2023-05-04 |
| Can LLM Already Serve as A Database Interface? A BIg Bench for Large-Scale Database Grounded Text-to-SQLs | ✓ Link | 33.04 | 27.38 | | Palm-2 (Baseline) | 2023-05-04 |
| MSc-SQL: Multi-Sample Critiquing Small Language Models For Text-To-SQL Translation | ✓ Link | | 65.6 | | MSc-SQL | 2024-10-16 |
| []() | | | 64.62 | | SFT CodeS-15B + SQLFixAgent | |
| Knowledge-to-SQL: Enhancing SQL Generation with Data Expert LLM | ✓ Link | | 48.92 | | DELLM + MAC-SQL | 2024-02-18 |
| Can LLM Already Serve as A Database Interface? A BIg Bench for Large-Scale Database Grounded Text-to-SQLs | ✓ Link | | | 92.96 | Human Performance | 2023-05-04 |