Text-to-Image Retrieval on ShareGPT4V
97.9R@1LamRA
Evaluation results
| Method | Links | |
|---|---|---|
| LamRAreranking=pointwise/listwise2024.12 | 97.9 | |
| ObjEmbedParameters=4B2026.02 | 97.7 | |
| ObjEmbedParameters=2B2026.02 | 97.3 | |
| FG-CLIPBackbone=ViT-L/14, Evaluation Protocol=Zero-shot2025.10 | 96.8 | |
| FG-CLIP2Backbone=ViT-So/162026.02 | 96.7 | |
| Qwen3-VL-EmbeddingParameters=2B2026.02 | 96.7 | |
| FG-CLIP 2Backbone=ViT-So/16, Evaluation Protocol=Zero-shot2025.10 | 96.7 | |
| FG-CLIP 2Backbone=ViT-L/16, Evaluation Protocol=Zero-shot2025.10 | 96.6 | |
| LamRAreranking=pointwise/listwise2024.12 | 96.5 | |
| Long-CLIP + GMAILalignment=GMAIL2026.02 | 96.1 | |
| Long-CLIP-L2024.12 | 95.8 | |
| Long-CLIP-L2024.12 | 95.6 | |
| Long-CLIPBackbone=ViT-L/14, Evaluation Protocol=Zero-shot2025.10 | 95.6 | |
| FG-CLIP 2Backbone=ViT-B/16, Evaluation Protocol=Zero-shot2025.10 | 95.4 | |
| FG-CLIPBackbone=ViT-B/16, Evaluation Protocol=Zero-shot2025.10 | 94.9 | |
| EVA-CLIPparameters=18B2024.12 | 94 | |
| UME-R1Parameters=2B2026.02 | 93.9 | |
| Long-CLIPBackbone=ViT-B/16, Evaluation Protocol=Zero-shot2025.10 | 93.4 | |
| LamRA-Ret2024.12 | 93.3 | |
| Long-CLIP2026.02 | 93.3 | |
| EVA-CLIPparameters=8B2024.12 | 93.1 | |
| VLM2Vec-V2Parameters=2B2026.02 | 92.4 | |
| GMEParameters=7B2026.02 | 92.2 | |
| EVA-CLIPparameters=18B2024.12 | 92.1 | |
| GMEParameters=2B2026.02 | 91.7 | |
| EVA-CLIPparameters=8B2024.12 | 91.2 | |
| EVA-CLIPBackbone=ViT-L/142026.02 | 89.4 | |
| EVA-CLIPBackbone=ViT-L/14, Evaluation Protocol=Zero-shot2025.10 | 89.4 | |
| MetaCLIP2Backbone=ViT-H/142026.02 | 89.2 | |
| Meta CLIP 2Backbone=ViT-H/14, Evaluation Protocol=Zero-shot2025.10 | 89.2 | |
| LamRA-Ret2024.12 | 88.1 | |
| E5-V2024.12 | 86.7 | |
| CLIP + GMAILalignment=GMAIL2026.02 | 86.7 | |
| UniIR-CLIP2024.12 | 85.8 | |
| MagicLens-L2024.12 | 85.5 | |
| EVA-CLIPBackbone=ViT-B/16, Evaluation Protocol=Zero-shot2025.10 | 85.5 | |
| SigLIP 2Backbone=ViT-L/16, Evaluation Protocol=Zero-shot2025.10 | 84.6 | |
| UniIR-CLIP2024.12 | 84.1 | |
| CLIP-L2024.12 | 84 | |
| E5-V2024.12 | 84 | |
| CLIPBackbone=ViT-L/142026.02 | 83.6 | |
| CLIPBackbone=ViT-L/14, Evaluation Protocol=Zero-shot2025.10 | 83.6 | |
| FineCLIPBackbone=ViT-L/14, Evaluation Protocol=Zero-shot2025.10 | 82.7 | |
| CLIP-L2024.12 | 81.8 | |
| CLIP2026.02 | 79.6 | |
| CLIPBackbone=ViT-B/16, Evaluation Protocol=Zero-shot2025.10 | 79.6 | |
| SigLIP2Backbone=ViT-So/162026.02 | 79.5 | |
| SigLIP 2Backbone=ViT-So/16, Evaluation Protocol=Zero-shot2025.10 | 79.5 | |
| VLM2VecParameters=2B2026.02 | 74.3 | |
| FineCLIPBackbone=ViT-B/16, Evaluation Protocol=Zero-shot2025.10 | 73.3 | |
| SigLIP 2Backbone=ViT-B/16, Evaluation Protocol=Zero-shot2025.10 | 67.9 | |
| MagicLens-L2024.12 | 60.9 |