Chart Question Answering on ChartQA
95.9AccuracyQwen3-VL-8B-Masters
Evaluation results
| Method | Links | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Qwen3-VL-8B-MastersSize=8B2025.12 | 95.9 | — | — | — | — | — | — | — | — | — | |
| MastersBase Model=Qwen3-VL-8B2025.12 | 95.9 | — | — | — | — | — | — | — | — | — | |
| Qwen2.5-VL-7B-MastersSize=7B2025.12 | 95.6 | — | — | — | — | — | — | — | — | — | |
| Qwen2.5-VL-3B-MastersSize=3B2025.12 | 95.6 | — | — | — | — | — | — | — | — | — | |
| MastersBase Model=Qwen2.5-VL-7B2025.12 | 95.6 | — | — | — | — | — | — | — | — | — | |
| InternVL3.5-8B-MastersSize=8B2025.12 | 95.1 | — | — | — | — | — | — | — | — | — | |
| MastersBase Model=InternVL3.5-8B2025.12 | 95.1 | — | — | — | — | — | — | — | — | — | |
| InternVL3.5-4B-MastersSize=4B2025.12 | 94.9 | — | — | — | — | — | — | — | — | — | |
| InternVL3-8B-MastersSize=8B2025.12 | 94.8 | — | — | — | — | — | — | — | — | — | |
| InternVL3.5-2B-MastersSize=2B2025.12 | 94.8 | — | — | — | — | — | — | — | — | — | |
| MastersBase Model=InternVL3-8B2025.12 | 94.8 | — | — | — | — | — | — | — | — | — | |
| Qwen3-VL-4B-MastersSize=4B2025.12 | 94.7 | — | — | — | — | — | — | — | — | — | |
| Qwen3-VL-2B-MastersSize=2B2025.12 | 94.1 | — | — | — | — | — | — | — | — | — | |
| InternVL2.5-8B-GenRecalTeacher VLM=InternVL2.5-78B2025.06 | 93.6 | — | — | — | — | — | — | — | — | — | |
| InternVL3-2B-MastersSize=2B2025.12 | 93.5 | — | — | — | — | — | — | — | — | — | |
| Claude-3.7-SonnetVariant=Sonnet 3.72025.12 | 92.2 | — | — | — | — | — | — | — | — | — | |
| MiMo-VL-8BSize=8B2025.12 | 91.7 | — | — | — | — | — | — | — | — | — | |
| Qwen2.5vl-InstructModel Scale=7B, Training Protocol=PSO, Category=Open-Source Reasoning MLLMs2025.12 | 91.2 | — | — | — | — | — | — | — | — | — | |
| RLR³Training Source=ViRL2026.05 | 91.1 | — | — | — | — | — | — | — | — | — | |
| Chart-R1Model size=7B2025.12 | 91 | — | — | — | — | — | — | — | — | — | |
| RLVRTraining Source=DeepVision2026.05 | 91 | — | — | — | — | — | — | — | — | — | |
| Claude-3.5-SonnetVariant=Sonnet 3.52025.12 | 90.8 | — | — | — | — | — | — | — | — | — | |
| Claude-3.5-Sonnet2025.06 | 90.8 | — | — | — | — | — | — | — | — | — | |
| Claude-3.5-SonnetSize=Closed2026.03 | 90.8 | — | — | — | — | — | — | — | — | — | |
| Claude-3.5-sonnet2025.02 | 90.8 | — | — | — | — | — | — | — | — | — | |
| RLVRTraining Source=ViRL2026.05 | 90.6 | — | — | — | — | — | — | — | — | — | |
| RLVRTraining Source=OpenMMR2026.05 | 90.6 | — | — | — | — | — | — | — | — | — | |
| RLR³Training Source=OpenMMR2026.05 | 90.6 | — | — | — | — | — | — | — | — | — | |
| InternVL2.5-8B-GenRecalTeacher VLM=InternVL2-76B2025.06 | 90.4 | — | — | — | — | — | — | — | — | — | |
| Qwen3-VLSize=235B(A22B)2026.03 | 90.3 | — | — | — | — | — | — | — | — | — | |
| RLR³Training Source=DeepVision2026.05 | 90.3 | — | — | — | — | — | — | — | — | — | |
| Qwen3-VL-32BSize=32B2025.12 | 89.8 | — | — | — | — | — | — | — | — | — | |
| InternVL3-78BSize=78B2025.12 | 89.7 | — | — | — | — | — | — | — | — | — | |
| InternVL3.5Size=78B2026.03 | 89.7 | — | — | — | — | — | — | — | — | — | |
| Qwen3-VLSize=8B2026.03 | 89.6 | — | — | — | — | — | — | — | — | — | |
| GPT52025.02 | 89.6 | — | — | — | — | — | — | — | — | — | |
| Qwen2.5-VL-72BSize=72B2025.12 | 89.5 | — | — | — | — | — | — | — | — | — | |
| Qwen2.5-VLSize=72B2026.03 | 89.5 | — | — | — | — | — | — | — | — | — | |
| Official thinkingSource=Qwen3-VL2026.05 | 89.4 | — | — | — | — | — | — | — | — | — | |
| Qwen2.5-VL-7B-InstructParameters=7B, Training Strategy=SFT+GRPO2025.05 | 89.2 | — | — | — | — | — | — | — | — | — | |
| Qwen2.5vl-InstructModel Scale=7B, Training Protocol=SFT + GRPO, Category=Open-Source Reasoning MLLMs2025.12 | 89.1 | — | — | — | — | — | — | — | — | — | |
| Qwen2.5-VL-7B-Instruct + GRPO + KAWHITraining Method=GRPO + KAWHI2026.03 | 89.1 | — | — | — | — | — | — | — | — | — | |
| SPOT-EBase Model=Qwen3-VL-8B, Source Type=Open-Source, Test-time Adaptation=SPOT-E2026.06 | 89 | — | — | — | — | — | — | — | — | — | |
| START-RLModel size=7B2025.12 | 88.8 | — | — | — | — | — | — | — | — | — | |
| InternVL3.5-38BSize=38B2025.12 | 88.8 | — | — | — | — | — | — | — | — | — | |
| InternVL3.5Size=38B2026.03 | 88.8 | — | — | — | — | — | — | — | — | — | |
| Base instruct2026.05 | 88.6 | — | — | — | — | — | — | — | — | — | |
| SophiaVL-R1-7BParameters=7B2025.05 | 88.5 | — | — | — | — | — | — | — | — | — | |
| SPOT-EBase Model=Qwen2.5-VL-7B, Source Type=Open-Source, Test-time Adaptation=SPOT-E2026.06 | 88.5 | — | — | — | — | — | — | — | — | — | |
| Qwen3-VL-8BSize=8B2025.12 | 88.4 | — | — | — | — | — | — | — | — | — | |
| InternVL2-76B2025.06 | 88.4 | — | — | — | — | — | — | — | — | — | |
| Qwen2-VL-72B2025.06 | 88.3 | — | — | — | — | — | — | — | — | — | |
| InternVL2.5-78B2025.06 | 88.3 | — | — | — | — | — | — | — | — | — | |
| Qwen2.5-VL-7B + ECRDParameters=7B, Enhancement=ECRD2026.02 | 88.3 | — | — | — | — | — | — | — | — | — | |
| Qwen2.5-VL-7B-Instruct + GRPOTraining Method=GRPO2026.03 | 88.3 | — | — | — | — | — | — | — | — | — | |
| GPT5-mini2025.02 | 88.2 | — | — | — | — | — | — | — | — | — | |
| SPOT-EBase Model=GPT-4o, Source Type=Closed-Source, Test-time Adaptation=SPOT-E2026.06 | 88.2 | — | — | — | — | — | — | — | — | — | |
| SPOT-EBase Model=InternVL3-8B, Source Type=Open-Source, Test-time Adaptation=SPOT-E2026.06 | 88.1 | — | — | — | — | — | — | — | — | — | |
| Qwen3-VL-8BBase Model=Qwen3-VL-8B, Source Type=Open-Source, Test-time Adaptation=None2026.06 | 88 | — | — | — | — | — | — | — | — | — | |
| ERNIE 5.0-BaseModel type=pre-trained2026.02 | 87.68 | — | — | — | — | — | — | — | — | — | |
| START-SFTModel size=7B2025.12 | 87.6 | — | — | — | — | — | — | — | — | — | |
| InternVL2.5-8B-GenRecalTeacher VLM=Qwen2-VL-72B2025.06 | 87.5 | — | — | — | — | — | — | — | — | — | |
| Qwen3-VL-4BSize=4B2025.12 | 87.4 | — | — | — | — | — | — | — | — | — | |
| Qwen2.5-VL-7B-InstructModel Size=7B2026.03 | 87.4 | — | — | — | — | — | — | — | — | — | |
| MMRPTModel Scale=Qwen2.5-VL-7B-Instruct, Evaluation Protocol=Zero-shot2025.12 | 87.33 | — | — | — | — | — | — | — | — | — | |
| Phantom-3.8BModel Size=3.8B2024.09 | 87.3 | — | — | — | — | — | — | — | — | — | |
| Qwen2.5-VLModel size=7B2025.12 | 87.3 | — | — | — | — | — | — | — | — | — | |
| Qwen2.5-VL-7BSize=7B2025.12 | 87.3 | — | — | — | — | — | — | — | — | — | |
| Molmo-72BSize=72B2025.12 | 87.3 | — | — | — | — | — | — | — | — | — | |
| Qwen2.5-VLInput=Any Res., RoPE=M-RoPE, Backbone=Dense, Parameter Scale=8B2025.10 | 87.3 | — | — | — | — | — | — | — | — | — | |
| Molmo-72B2025.06 | 87.3 | — | — | — | — | — | — | — | — | — | |
| MolmoSize=72B2026.03 | 87.3 | — | — | — | — | — | — | — | — | — | |
| Qwen2.5-VLSize=7B2026.03 | 87.3 | — | — | — | — | — | — | — | — | — | |
| Qwen2.5-VL-7BBase Model=Qwen2.5-VL-7B, Source Type=Open-Source, Test-time Adaptation=None2026.06 | 87.3 | — | — | — | — | — | — | — | — | — | |
| BaseModel Scale=Qwen2.5-VL-7B-Instruct, Evaluation Protocol=Zero-shot2025.12 | 87.25 | — | — | — | — | — | — | — | — | — | |
| Gemini-1.5-ProVariant=Pro 1.52025.12 | 87.2 | — | — | — | — | — | — | — | — | — | |
| Gemini-1.5-Pro2025.06 | 87.2 | — | — | — | — | — | — | — | — | — | |
| Qwen2.5-VL-7B-InstructParameters=7B, Training Strategy=GRPO2025.05 | 87.2 | — | — | — | — | — | — | — | — | — | |
| Q-MaskSize=3B2026.03 | 87.2 | — | — | — | — | — | — | — | — | — | |
| LLaVA-OneVision-1.5-4BSize=4B2025.12 | 87.1 | — | — | — | — | — | — | — | — | — | |
| Phantom-1.8BModel Size=1.8B2024.09 | 87 | — | — | — | — | — | — | — | — | — | |
| Qwen2.5-VL-7BBackbone=Qwen2.5-VL-7B2026.03 | 87 | — | — | — | — | — | — | — | — | — | |
| MiniCPM-o2.6-8BSize=8B2025.12 | 86.9 | — | — | — | — | — | — | — | — | — | |
| Qwen2.5-VL-7B + supervisorParameters=7B, Enhancement=supervisor2026.02 | 86.8 | — | — | — | — | — | — | — | — | — | |
| InternVL3.5-14B +FINER-TuningBackbone=InternVL3.5-14B, Training Strategy=FINER-Tuning2026.03 | 86.8 | — | — | — | — | — | — | — | — | — | |
| Official instructSource=Qwen3-VL2026.05 | 86.8 | — | — | — | — | — | — | — | — | — | |
| InternVL3.5-8BSize=8B2025.12 | 86.7 | — | — | — | — | — | — | — | — | — | |
| InternVL3.5-8BBackbone=InternVL3.5-8B2026.03 | 86.7 | — | — | — | — | — | — | — | — | — | |
| InternVL3.5-8B +FINER-TuningBackbone=InternVL3.5-8B, Training Strategy=FINER-Tuning2026.03 | 86.7 | — | — | — | — | — | — | — | — | — | |
| InternVL3.5Size=8B2026.03 | 86.7 | — | — | — | — | — | — | — | — | — | |
| GPT-4oBase Model=GPT-4o, Source Type=Closed-Source, Test-time Adaptation=None2026.06 | 86.7 | — | — | — | — | — | — | — | — | — | |
| InternVL3-8BSize=8B2025.12 | 86.6 | — | — | — | — | — | — | — | — | — | |
| GLM-4.5V2025.12 | 86.6 | — | — | — | — | — | — | — | — | — | |
| InternVL3Input=Tile-wise, RoPE=1D-RoPE, Backbone=Dense, Parameter Scale=8B2025.10 | 86.6 | — | — | — | — | — | — | — | — | — | |
| GLM-4.5VSize=106B(A12B)2026.03 | 86.6 | — | — | — | — | — | — | — | — | — | |
| InternVL3.5Size=20B2026.03 | 86.6 | — | — | — | — | — | — | — | — | — | |
| InternVL3-8BBase Model=InternVL3-8B, Source Type=Open-Source, Test-time Adaptation=None2026.06 | 86.6 | — | — | — | — | — | — | — | — | — | |
| LLaVA-OneVision-1.5-8BSize=8B2025.12 | 86.5 | — | — | — | — | — | — | — | — | — | |
| TokenVL-8BSize=8B2026.03 | 86.5 | — | — | — | — | — | — | — | — | — | |
| Qwen2.5-VL-7B +FINER-TuningBackbone=Qwen2.5-VL-7B, Training Strategy=FINER-Tuning2026.03 | 86.4 | — | — | — | — | — | — | — | — | — |