Share your thoughts, 1 month free Claude Pro on usSee more

Table Understanding on TableBench

88.54Fact Score

Qwen3-32B

Updated 1mo ago

Evaluation Results

Method	Links
Qwen3-32B 2026.06		88.54	84.13	48.4	69.98
Claude-4.5-Sonnet 2026.06		87.5	85.39	53.64	72.61
VeriGraph-8B 2026.06		87.23	83.63	57.78	73.58
Claude-4.5-Opus 2026.06		86.15	86.46	57.43	74.4
GPT-5.2 2026.06		85.42	77.83	48.4	66.63
Claude-4.5-Opus 2026.06		84.38	87.41	55.69	74.04
Qwen3-30B-A3B 2026.06		84.38	85.89	46.06	69.38
Gemini-2.5-Pro 2026.06		82.29	82.62	45.19	67.22
DeepAnalyze 2026.06		80.21	71.92	50.4	67.51
Qwen3-32B 2026.06		78.12	79.09	50.73	67.34
GPT-5.4 2026.06		77.08	84.63	62.1	74.52
QwQ-32B 2026.06		77.08	70.03	51.31	63.16
GPT-5.2 2026.06		75	79.09	52.48	67.7
DataMind 2026.06		73.96	82.12	50.73	68.94
Claude-4.5-Sonnet 2026.06		72.92	72.07	43.01	62.52
Qwen3-30B-A3B 2026.06		68.75	72.54	51.6	63.52
Qwen3-Coder-30B 2026.06		65.62	60.15	35.86	50.78
Qwen3-8B 2026.06		60.42	72.54	52.19	62.8