Multimodal Dense Retrieval on MMEB V2 benchmark leaderboard · SOTA2 Research

Multimodal Dense Retrieval on MMEB V2

80.15Image Recall@1Qwen3-VL-Embed-8B

Evaluation Results

Method	Links
Qwen3-VL-Embed-8B2026.05		80.15	87.92	70.57	89.12	73.21	60.43	79.34
Qwen3-VL-Embed-8B2026.05		80.09	87.29	69.35	88.78	73.27	59.01	78.83
Qwen3-VL-Embed-2B2026.05		75.33	85.52	66.61	86.87	69.9	56.05	75.77
Qwen3-VL-Embed-2B2026.05		74.91	84.46	65.38	86.19	69.37	54.04	74.87
GME-7B2026.05		73.57	90.12	61.04	87.05	69.43	29.38	72.56
GME-7B2026.05		73.09	90.01	60.43	86.2	69.22	29.92	72.26
GME-2B2026.05		71.05	87.97	57.23	84.18	67.82	26.5	70
GME-2B2026.05		70.13	87.61	53.76	82.49	66.93	26.73	69
VLM2Vec-V2.02026.05		69.95	80.23	51.08	82.76	64	30.03	67.04
VLM2Vec-V2.02026.05		69.5	75.34	47.28	79.43	62.04	28.66	64.5