Text-to-SQL on Spider
98.65Exec Acc (All)SQLCoder
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| SQLCoderBest alpha=0.41, Strategy=Best Pairwise2026.05 | 98.65 | — | — | — | — | — | |
| Llama 3.1Best alpha=0.32, Strategy=Best Pairwise2026.05 | 98.38 | — | — | — | — | — | |
| Gemma 3Best alpha=0.41, Strategy=Best Pairwise2026.05 | 98.38 | — | — | — | — | — | |
| SQLCoderBest alpha=0.41, Strategy=Best kNN2026.05 | 98.38 | — | — | — | — | — | |
| Gemma 3Best alpha=0.41, Strategy=Best kNN2026.05 | 98.38 | — | — | — | — | — | |
| Qwen 3Best alpha=0.32, Strategy=Best Pairwise2026.05 | 98.11 | — | — | — | — | — | |
| Llama 3.1Best alpha=0.32, Strategy=Best kNN2026.05 | 98.11 | — | — | — | — | — | |
| Qwen 3Best alpha=0.32, Strategy=Best kNN2026.05 | 95.14 | — | — | — | — | — | |
| IQuest-Coder-V1-40B-InstructModel Size Category=20B+2026.03 | 92.2 | — | — | — | — | — | |
| CHASE-SQL + Gemini 1.5Model Size=Closed-Source (very large), Use specific training set?=YES, Support Multi-Turn Text-to-SQL?=NO2025.10 | 87.6 | — | — | — | — | — | |
| Gemini-3-Flash-previewModel Size Category=Closed-APIs Models2026.03 | 87.2 | — | — | — | — | — | |
| Gemini-3-Pro-previewModel Size Category=Closed-APIs Models2026.03 | 87 | — | — | — | — | — | |
| SQL-O1Backbone=XiYanSQL, Size=7B, tuning=true2026.02 | 86.54 | — | — | — | — | — | |
| DAIL-SQL + GPT-4Model Size=Closed-Source (very large), Use specific training set?=YES, Support Multi-Turn Text-to-SQL?=NO2025.10 | 86.2 | — | — | — | — | — | |
| LinkAlignBackbone=XiYanSQL, Size=7B2026.02 | 85.89 | — | — | — | — | — | |
| SQL-O1Backbone=Seed-Coder, Size=8B, tuning=true2026.02 | 85.25 | — | — | — | — | — | |
| SQL-O1Backbone=Qwen2.5-Coder, Size=7B, tuning=true2026.02 | 85.1 | — | — | — | — | — | |
| DTS-SQLBackbone=XiYanSQL, Size=7B2026.02 | 85.09 | — | — | — | — | — | |
| DTS-SQLBackbone=OmniSQL, Size=7B2026.02 | 84.99 | — | — | — | — | — | |
| Alpha-SQLBackbone=XiYanSQL, Size=7B2026.02 | 84.93 | — | — | — | — | — | |
| Alpha-SQLBackbone=Seed-Coder, Size=8B2026.02 | 84.37 | — | — | — | — | — | |
| Qwen3-4B + Long-Horizon Training (Ours)Model Size=4B, Use specific training set?=NO, Support Multi-Turn Text-to-SQL?=YES2025.10 | 84.2 | — | — | — | — | — | |
| DTS-SQLBackbone=Qwen2.5-Coder, Size=7B2026.02 | 84.12 | — | — | — | — | — | |
| Alpha-SQLBackbone=Qwen2.5-Coder, Size=7B2026.02 | 84 | — | — | — | — | — | |
| IQuest-Coder-V1-40B-Loop-InstructModel Size Category=20B+2026.03 | 84 | — | — | — | — | — | |
| Qwen2.5-Coder-32B-InstructModel Size Category=20B+2026.03 | 83.9 | — | — | — | — | — | |
| IESRBackbone=Qwen2.5-Coder, Size=7B2026.02 | 83.45 | — | — | — | — | — | |
| IESRBackbone=XiYanSQL, Size=7B2026.02 | 83.36 | — | — | — | — | — | |
| CHESS-SQLBackbone=XiYanSQL, Size=7B2026.02 | 83.3 | — | — | — | — | — | |
| DTS-SQLBackbone=Seed-Coder, Size=8B2026.02 | 83.21 | — | — | — | — | — | |
| DAIL-sqlBackbone=Seed-Coder, Size=8B2026.02 | 82.98 | — | — | — | — | — | |
| DAIL-sqlBackbone=XiYanSQL, Size=7B2026.02 | 82.56 | — | — | — | — | — | |
| LLaDA2.0-flash2026.02 | 82.49 | — | — | — | — | 4.42 | |
| LinkAlignBackbone=OmniSQL, Size=7B2026.02 | 82.32 | — | — | — | — | — | |
| LinkAlignBackbone=Seed-Coder, Size=8B2026.02 | 82.12 | — | — | — | — | — | |
| Qwen3-30B-A3B-Inst-25072026.02 | 81.79 | — | — | — | — | 1 | |
| DAIL-sqlBackbone=OmniSQL, Size=7B2026.02 | 81.56 | — | — | — | — | — | |
| CHESS-SQLBackbone=Seed-Coder, Size=8B2026.02 | 81.31 | — | — | — | — | — | |
| Qwen2.5-Coder-14B-InstructModel Size Category=13B+2026.03 | 81.3 | — | — | — | — | — | |
| Qwen3-Coder-480B-A35B-InstructModel Size Category=20B+2026.03 | 81.2 | — | — | — | — | — | |
| SQL-O1Backbone=OmniSQL, Size=7B, tuning=true2026.02 | 81.12 | — | — | — | — | — | |
| Qwen3-235B-A22B-Instruct-2507Model Size Category=20B+2026.03 | 81.1 | — | — | — | — | — | |
| Kimi-K2-Instruct-0905Model Size Category=20B+2026.03 | 81.1 | — | — | — | — | — | |
| LLaDA2.1-flashInference Mode=Q Mode2026.02 | 81.04 | — | — | — | — | 5.7 | |
| Qwen3-Coder-30B-A3B-InstructModel Size Category=13B+2026.03 | 80.9 | — | — | — | — | — | |
| Ling-flash-2.02026.02 | 80.58 | — | — | — | — | 1 | |
| DAIL-sqlBackbone=Qwen2.5-Coder, Size=7B2026.02 | 80.31 | — | — | — | — | — | |
| CHESS-SQLBackbone=Qwen2.5-Coder, Size=7B2026.02 | 80.31 | — | — | — | — | — | |
| LinkAlignBackbone=Qwen2.5-Coder, Size=7B2026.02 | 80.11 | — | — | — | — | — | |
| Claude-Sonnet-4.5Model Size Category=Closed-APIs Models2026.03 | 80.1 | — | — | — | — | — | |
| IESRBackbone=OmniSQL, Size=7B2026.02 | 80.02 | — | — | — | — | — | |
| Qwen2.5-Coder-7B-InstructModel Size Category=6B+2026.03 | 79.8 | — | — | — | — | — | |
| Alpha-SQLBackbone=OmniSQL, Size=7B2026.02 | 79.34 | — | — | — | — | — | |
| Din-sqlBackbone=XiYanSQL, Size=7B2026.02 | 79.3 | — | — | — | — | — | |
| IESRBackbone=Seed-Coder, Size=8B2026.02 | 79.22 | — | — | — | — | — | |
| LLaDA2.1-flashInference Mode=S Mode2026.02 | 79.18 | — | — | — | — | 8.74 | |
| IQuest-Coder-V1-40B-ThinkingModel Size Category=20B+2026.03 | 78.1 | — | — | — | — | — | |
| Din-sqlBackbone=Seed-Coder, Size=8B2026.02 | 78.01 | — | — | — | — | — | |
| RESDSQL-3B + NatSQModel Size=3B, Use specific training set?=YES, Support Multi-Turn Text-to-SQL?=NO2025.10 | 78 | — | — | — | — | — | |
| DeepSeek-v3.2Model Size Category=20B+2026.03 | 77.9 | — | — | — | — | — | |
| IQuest-Coder-V1-40B-Loop-ThinkingModel Size Category=20B+2026.03 | 77.8 | — | — | — | — | — | |
| PV-SQLBase LLM=GPT-4o2026.04 | 77.66 | — | — | — | — | — | |
| KAT-DevModel Size Category=20B+2026.03 | 77.6 | — | — | — | — | — | |
| GPT-5.1Model Size Category=Closed-APIs Models2026.03 | 77.6 | — | — | — | — | — | |
| LLaDA2.1-minimode=Q Mode2026.02 | 77.55 | — | — | — | — | 5.48 | |
| Din-sqlBackbone=Qwen2.5-Coder, Size=7B2026.02 | 77.19 | — | — | — | — | — | |
| CHESS-SQLBackbone=OmniSQL, Size=7B2026.02 | 77.02 | — | — | — | — | — | |
| Din-sqlBackbone=OmniSQL, Size=7B2026.02 | 76.8 | — | — | — | — | — | |
| LLaDA2.0-mini2026.02 | 76.76 | — | — | — | — | 3.93 | |
| Ling-mini-2.02026.02 | 76.43 | — | — | — | — | — | |
| Claude-Opus-4.5Model Size Category=Closed-APIs Models2026.03 | 76 | — | — | — | — | — | |
| IQuest-Coder-V1-14B-InstructModel Size Category=13B+2026.03 | 75.8 | — | — | — | — | — | |
| LLaDA2.1-minimode=S Mode2026.02 | 75.78 | — | — | — | — | 7.98 | |
| E-SQLBase LLM=GPT-4o2026.04 | 75.63 | — | — | — | — | — | |
| TA-SQLBase LLM=GPT-4o2026.04 | 74.66 | — | — | — | — | — | |
| TS-SQLBase LLM=GPT-4o2026.04 | 74.18 | — | — | — | — | — | |
| Qwen3-14BModel Size=14B, Use specific training set?=NO, Support Multi-Turn Text-to-SQL?=YES2025.10 | 72.9 | — | — | — | — | — | |
| Qwen3-8Bno think=true2026.02 | 72.8 | — | — | — | — | — | |
| MAC-SQLBase LLM=GPT-4o2026.04 | 72.73 | — | — | — | — | — | |
| XiYan-SQLBase LLM=GPT-4o2026.04 | 72.73 | — | — | — | — | — | |
| Seed-Coder-8B-InstructModel Size Category=6B+2026.03 | 72.7 | — | — | — | — | — | |
| DeepSeek-Coder-V2-Lite-InstructModel Size Category=6B+2026.03 | 72.4 | — | — | — | — | — | |
| DAIL-SQLBase LLM=GPT-4o2026.04 | 71.18 | — | — | — | — | — | |
| GPT-4o2026.04 | 71.08 | — | — | — | — | — | |
| Qwen3-4BModel Size=4B, Use specific training set?=NO, Support Multi-Turn Text-to-SQL?=YES2025.10 | 70.7 | — | — | — | — | — | |
| Table-SpecialistBase Model=GPT-4, Training Strategy=Table-Specialist2024.10 | 70.4 | — | — | — | — | — | |
| GPT-4Base Model=GPT-4, Training Strategy=Vanilla2024.10 | 69.4 | — | — | — | — | — | |
| IQuest-Coder-V1-14B-ThinkingModel Size Category=13B+2026.03 | 69.3 | — | — | — | — | — | |
| Cot-sqlBackbone=XiYanSQL, Size=7B2026.02 | 69.12 | — | — | — | — | — | |
| DIN-SQLBase LLM=GPT-4o2026.04 | 69.05 | — | — | — | — | — | |
| Table-SpecialistBase Model=Llama3.1-8B, Training Strategy=Table-Specialist2024.10 | 69 | — | — | — | — | — | |
| Table-SpecialistBase Model=GPT-3.5, Training Strategy=Table-Specialist2024.10 | 68.4 | — | — | — | — | — | |
| Cot-sqlBackbone=OmniSQL, Size=7B2026.02 | 68.32 | — | — | — | — | — | |
| IQuest-Coder-V1-7B-InstructModel Size Category=6B+2026.03 | 67.7 | — | — | — | — | — | |
| Llama3.1-8BBase Model=Llama3.1-8B, Training Strategy=Vanilla2024.10 | 67 | — | — | — | — | — | |
| OracleRetrieval category=Oracle2026.04 | 66.6 | — | — | — | — | — | |
| Cot-sqlBackbone=Seed-Coder, Size=8B2026.02 | 66.23 | — | — | — | — | — | |
| Full Fine-tuning (All)Model Size=Mamba 1.4B, Params (%)=100.00, Type=None2025.03 | 66.2 | 84.3 | 69.5 | 53.4 | 43.4 | — | |
| Cot-sqlBackbone=Qwen2.5-Coder, Size=7B2026.02 | 65.32 | — | — | — | — | — | |
| GPT-3.5Base Model=GPT-3.5, Training Strategy=Vanilla2024.10 | 65 | — | — | — | — | — |