Table Question Answering on TAT-QA
75.88AccuracyQwen3-VL-8B + TABQAWORLD
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Qwen3-VL-8B + TABQAWORLDStrategy=Training-free Agent2026.04 | 75.88 | — | |
| Qwen2.5-VL-7B + TABQAWORLDStrategy=Training-free Agent2026.04 | 71.36 | — | |
| Qwen2.5-VL-Instruct + Vanilla SFTScale=7B, Modality=Image & Text2025.02 | 62.95 | 62.15 | |
| Qwen2.5-VL-Instruct + HIPPOScale=7B, Modality=Image & Text2025.02 | 62.82 | 62.9 | |
| TableDART (TG2-7B+Ovis2-8B)Strategy=Dynamic Adaptive Routing2026.04 | 62.05 | — | |
| MiniCPM-V-2.6 + HIPPOScale=8B, Modality=Image & Text2025.02 | 60.75 | 65.46 | |
| HIPPO-8BModality=Multimodal2026.04 | 60.75 | — | |
| Qwen3-VL-8BModality=Image2026.04 | 59.8 | — | |
| TableDART (TG2-7B+Qwen2.5-VL-7B)Strategy=Dynamic Adaptive Routing2026.04 | 59.07 | — | |
| MiniCPM-V-2.6 + Vanilla SFTScale=8B, Modality=Image & Text2025.02 | 58.91 | 61.93 | |
| Qwen2.5-VL-InstructScale=7B, Modality=Image & Text2025.02 | 56.73 | 58.58 | |
| MiniCPM-V-2.6Scale=8B, Modality=Image & Text2025.02 | 52.46 | 62.12 | |
| SynTab-LLaVA-7BModality=Image2026.04 | 51.94 | — | |
| Qwen2.5-VL-7BModality=Image2026.04 | 51.94 | — | |
| MiniCPM-V-2.6 (Image Only)Scale=8B, Modality=Image2025.02 | 51.55 | 60.42 | |
| MiniCPM-V-2.6-8BModality=Image2026.04 | 51.55 | — | |
| Qwen2.5-VL-Instruct (Image Only)Scale=7B, Modality=Image2025.02 | 50.9 | 56.85 | |
| TableGPT2-7B (Text-only Path)Strategy=Dynamic Adaptive Routing2026.04 | 50.39 | — | |
| Ovis2-8B (Image-only Path)Strategy=Dynamic Adaptive Routing2026.04 | 47.67 | — | |
| Table-LLaVA + HIPPOScale=7B, Modality=Image & Text2025.02 | 43.26 | 55.52 | |
| Gemini 2.0 FlashModality=Multimodal2026.04 | 35.62 | — | |
| Table-LLaVAScale=7B, Modality=Image & Text2025.02 | 27.84 | 50.98 | |
| GPT-4o (Image Only)Scale=N/A, Modality=Image2025.02 | 25.12 | 54.26 | |
| GPT-4oScale=N/A, Modality=Image & Text2025.02 | 21.24 | 55.58 | |
| TableLlama + MarkdownScale=7B, Modality=Text2025.02 | 19.04 | 37.86 | |
| TableLlama-7BModality=Text2026.04 | 19.04 | — | |
| Table-LLaVA (Image Only)Scale=13B, Modality=Image2025.02 | 15.67 | 38.09 | |
| Llama2Scale=7B, Modality=Text2025.02 | 13.73 | 17.53 | |
| Llama-2-7BModality=Text2026.04 | 13.73 | — | |
| Llama3-InstructScale=8B, Modality=Text2025.02 | 13.08 | 31.98 | |
| Llama3-Instruct-8BModality=Text2026.04 | 13.08 | — | |
| Table-LLaVA (Image Only)Scale=7B, Modality=Image2025.02 | 12.82 | 35.69 | |
| Table-LLaVA-7BModality=Image2026.04 | 12.82 | — | |
| MonkeyScale=7B, Modality=Image2025.02 | 12.31 | 14.16 | |
| Llama3.1-InstructScale=8B, Modality=Text2025.02 | 10.06 | 35.42 | |
| TabPediaScale=7B, Modality=Image2025.02 | 9.59 | 20.84 | |
| Qwen2.5-InstructScale=7B, Modality=Text2025.02 | 9.27 | 61.54 | |
| Vary-toyScale=1.8B, Modality=Image2025.02 | 8.81 | 5.77 | |
| Qwen2-72B-InstructScale=72B, Modality=Text2025.02 | 4.87 | 57.11 | |
| LLaVA v1.5Scale=7B, Modality=Image2025.02 | 2.97 | 9.82 | |
| mPLUG-Owl2Scale=7B, Modality=Image2025.02 | 0.39 | 7.76 | |
| InternLM-XComposerScale=7B, Modality=Image2025.02 | 0.26 | 0.77 | |
| MiniGPT-4Scale=7B, Modality=Image2025.02 | 0.13 | 0.28 | |
| Qwen-VLScale=7B, Modality=Image2025.02 | 0.13 | 0.83 | |
| mPLUG-OwlScale=7B, Modality=Image2025.02 | 0.13 | 3.16 | |
| TableLlama + OracleScale=7B, Modality=Text2025.02 | — | 55.33 |