Counting on CountBenchQA
93.84AccuracyFineViT-VL
Evaluation Results
| Method | Links | |
|---|---|---|
| FineViT-VLLanguage Backbone=Qwen3 1.7B LLM, Evaluation Framework=VLMEvalKit [69]2026.03 | 93.84 | |
| Molmo2-4BModel=Molmo2-4B2026.05 | 91.2 | |
| Qwen2.5-VL2026.01 | 89.1 | |
| ZAYA1-VL-8B-A1BModel=ZAYA1-VL-8B-A1B2026.05 | 88.1 | |
| PLM-3BModel=PLM-3B2026.05 | 88.1 | |
| Qwen3-VL-2BModel=Qwen3-VL-2B2026.05 | 87.9 | |
| Qwen3-VL-4BModel=Qwen3-VL-4B2026.05 | 87.3 | |
| Qwen3-VLLanguage Backbone=Qwen3 1.7B LLM, Evaluation Framework=VLMEvalKit [69]2026.03 | 86.65 | |
| DeepSeek-VL2-16B-A2.4BModel=DeepSeek-VL2-16B-A2.4B2026.05 | 86 | |
| Molmo2026.01 | 85.8 | |
| Qwen3.5-4BModel=Qwen3.5-4B2026.05 | 84.8 | |
| GPT-5-miniGrounded=true2025.12 | 84.7 | |
| GPT-5-miniGrounded=false2025.12 | 84.3 | |
| Qwen3.5-2BModel=Qwen3.5-2B2026.05 | 84.2 | |
| InternVL3.5-4BModel=InternVL3.5-4B2026.05 | 82.5 | |
| LLaVA-OneVision2026.01 | 82.3 | |
| InternVL3.5-20B-A4BModel=InternVL3.5-20B-A4B2026.05 | 82.1 | |
| Aquila-VLLanguage Backbone=Qwen2.5 1.5B LLM, Evaluation Framework=VLMEvalKit [69]2026.03 | 80.69 | |
| MolmoE-8B-A1BModel=MolmoE-8B-A1B2026.05 | 77.4 | |
| Qwen2.5-VL-3BModel=Qwen2.5-VL-3B2026.05 | 77 | |
| VALOR2025.12 | 75.9 | |
| VALORModel Category=Open-Source, Tool Use=true2025.12 | 75.9 | |
| LLaVA-AIMv2encoder=AIMv22026.01 | 73.9 | |
| LLaVA-AIMv2-2D-RoPEencoder=AIMv2, positional_embedding=2D-RoPE2026.01 | 71.9 | |
| Gemma3-4b-it2026.01 | 71.3 | |
| Intern3.5-VLLanguage Backbone=Qwen3 1.7B LLM, Evaluation Framework=VLMEvalKit [69]2026.03 | 70.22 | |
| InternVL3.5-2BModel=InternVL3.5-2B2026.05 | 70 | |
| SmolVLM22026.01 | 69.2 | |
| Qwen3-8BModel Category=Open-Source, Tool Use=true2025.12 | 68.6 | |
| Gemma-3-12BModel Category=Open-Source, Tool Use=true2025.12 | 67.8 | |
| GPT-4oModel Category=Proprietary, Tool Use=true2025.12 | 67.6 | |
| VALOR-RLModel Category=Open-Source, Tool Use=true2025.12 | 67.6 | |
| PaliGemma2026.01 | 67.4 | |
| o4-miniModel Category=Proprietary, Tool Use=true2025.12 | 66.8 | |
| Llama-3.2-11BModel Category=Open-Source, Tool Use=true2025.12 | 66.2 | |
| Gemini-2.5-FlashModel Category=Proprietary, Tool Use=true2025.12 | 65.6 | |
| Claude-3.5-HaikuModel Category=Proprietary, Tool Use=true2025.12 | 65.2 | |
| Gemini-2.0-FlashModel Category=Proprietary, Tool Use=true2025.12 | 62.1 | |
| LLaVA-SigLIPencoder=SigLIP2026.01 | 58.1 | |
| LLaVA-SigLIP2encoder=SigLIP22026.01 | 53.2 | |
| LLaVA-NeXT2026.01 | 51.5 | |
| LLaVA-SigLIP-2D-RoPEencoder=SigLIP, positional_embedding=2D-RoPE2026.01 | 48.3 | |
| LLaVA v1.52026.01 | 46.8 | |
| LLaVA-SigLIP2-2D-RoPEencoder=SigLIP2, positional_embedding=2D-RoPE2026.01 | 40.2 | |
| LLaVA-2D-RoPEpositional_embedding=2D-RoPE2026.01 | 29 |