Text-to-SQL on Synthesized Business Logic-Driven Dataset benchmark leaderboard · SOTA2 Research

Text-to-SQL on Synthesized Business Logic-Driven Dataset

90.24Single Execution AccuracyGPT-5

Evaluation Results

Method	Links
GPT-52026.01		90.24	70.97	80.95	76.19	80.7
Gemini-2.5-Pro2026.01		75.61	64.52	71.43	42.86	65.79
Claude-4.5-Sonnet2026.01		73.17	48.39	57.14	28.57	55.26
Qwen3-235B2026.01		70.73	29.03	47.62	28.57	47.37
Qwen3-30B2026.01		68.29	29.03	38.1	19.05	42.98
Gemini-3-Pro2026.01		65.85	67.74	76.19	33.33	62.28
DeepSeek-R12026.01		63.41	41.94	42.86	33.33	48.25
DeepSeek-v3.22026.01		58.54	25.81	38.1	28.57	40.35
Qwen3-Coder-30b2026.01		51.22	25.81	42.86	23.81	37.72