Context-Aware Image Retrieval on DISBench Overall
32EMClaude-Opus-4.5-20251101
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Claude-Opus-4.5-20251101Framework=ImageSeeker, Embedding Model=Qwen3-VL-Embedding-2B2026.02 | 32 | 55.5 | |
| Claude-Opus-4.5-20251101Framework=ImageSeeker, Embedding Model=Qwen3-VL-Embedding-8B2026.02 | 28.7 | 55 | |
| Gemini-3-Pro-PreviewFramework=ImageSeeker, Embedding Model=Qwen3-VL-Embedding-2B2026.02 | 25.4 | 47.9 | |
| Gemini-3-Pro-PreviewFramework=ImageSeeker, Embedding Model=Qwen3-VL-Embedding-8B2026.02 | 24.6 | 47.9 | |
| Claude-Sonnet-4.5-20250929Framework=ImageSeeker, Embedding Model=Qwen3-VL-Embedding-8B2026.02 | 22.1 | 43.8 | |
| Claude-Sonnet-4.5-20250929Framework=ImageSeeker, Embedding Model=Qwen3-VL-Embedding-2B2026.02 | 17.2 | 39.4 | |
| GPT-5.2Framework=ImageSeeker, Embedding Model=Qwen3-VL-Embedding-8B2026.02 | 13.1 | 30 | |
| Gemini-3-Flash-PreviewFramework=ImageSeeker, Embedding Model=Qwen3-VL-Embedding-8B2026.02 | 12.3 | 36.8 | |
| GLM-4.6VFramework=ImageSeeker, Embedding Model=Qwen3-VL-Embedding-2B2026.02 | 12.3 | 30.4 | |
| GPT-5.2Framework=ImageSeeker, Embedding Model=Qwen3-VL-Embedding-2B2026.02 | 11.5 | 35.1 | |
| Gemini-3-Flash-PreviewFramework=ImageSeeker, Embedding Model=Qwen3-VL-Embedding-2B2026.02 | 11.5 | 33.9 | |
| Qwen3-VL-235B-A22B-InstructFramework=ImageSeeker, Embedding Model=Qwen3-VL-Embedding-8B2026.02 | 11.5 | 27.2 | |
| Qwen3-VL-32B-InstructFramework=ImageSeeker, Embedding Model=Qwen3-VL-Embedding-2B2026.02 | 10.7 | 25.4 | |
| Qwen3-VL-235B-A22B-ThinkingFramework=ImageSeeker, Embedding Model=Qwen3-VL-Embedding-2B2026.02 | 9.8 | 22.8 | |
| Qwen3-VL-235B-A22B-InstructFramework=ImageSeeker, Embedding Model=Qwen3-VL-Embedding-2B2026.02 | 9 | 25.3 | |
| Qwen3-VL-235B-A22B-ThinkingFramework=ImageSeeker, Embedding Model=Qwen3-VL-Embedding-8B2026.02 | 8.2 | 20.4 | |
| Qwen3-VL-32B-InstructFramework=ImageSeeker, Embedding Model=Qwen3-VL-Embedding-8B2026.02 | 8.2 | 20.9 | |
| GLM-4.6VFramework=ImageSeeker, Embedding Model=Qwen3-VL-Embedding-8B2026.02 | 8.2 | 26.1 | |
| GPT-4oFramework=ImageSeeker, Embedding Model=Qwen3-VL-Embedding-2B2026.02 | 7.4 | 22.2 | |
| GPT-4oFramework=ImageSeeker, Embedding Model=Qwen3-VL-Embedding-8B2026.02 | 5.7 | 21.8 |