Reranking on MTEB multilingual v2 benchmark leaderboard · SOTA2 Research

Reranking on MTEB multilingual v2

85.13Allo ScoreQwen/Qwen3-Embedding-4B

Evaluation Results

Method
Qwen/Qwen3-Embedding-4B2026.06	85.13	72.28	67.27	65.61	11.3	88.89	65.08
Qwen/Qwen3-Embedding-8B2026.06	84.88	73.47	68.29	65.08	11.79	90.28	65.63
Querit/Querit-Reranker-4B2026.06	83.89	79.84	69.39	70.74	30.87	91.78	71.08
codefuse-ai/F2LLM-v2-14B2026.06	83.44	78.3	67.22	83.66	19.99	90.3	70.48
codefuse-ai/F2LLM-v2-8B2026.06	82.71	77.61	67.41	82.17	22.46	89.65	70.33
codefuse-ai/F2LLM-v2-4B2026.06	82.25	76.43	67.27	81.73	19.19	89.4	69.38
google/gemini-embedding-0012026.06	81.77	73.84	67.95	66.73	10.97	92.24	65.58
jinaai/jina-embeddings-v5-text-small2026.06	81.39	74.89	68.04	68.84	11.33	89.46	65.66
jinaai/jina-embeddings-v5-omni-small2026.06	81.39	74.89	68.04	68.84	11.33	89.46	65.66
Alibaba-NLP/gte-Qwen2-7B-instruct2026.06	81.1	74.13	67.8	70.06	12.5	87.73	65.55
codefuse-ai/F2LLM-v2-1.7B2026.06	80.11	75.33	66.96	75.94	15.62	89.08	67.17
Linq-AI-Research/Linq-Embed-Mistral2026.06	79.82	76.84	66.89	61.24	11.28	90.13	64.37
jinaai/jina-embeddings-v5-text-nano2026.06	79.67	73.68	67.63	67.02	10.95	88.83	64.63
jinaai/jina-embeddings-v5-omni-nano2026.06	79.67	73.68	67.63	67.02	10.95	88.83	64.63
Querit/Querit-Reranker-A0.4B2026.06	79.39	75.37	68.98	68.59	11.63	91.05	65.84
Salesforce/SFR-Embedding-Mistral2026.06	79.32	77.24	66.97	62.06	9.38	90.17	64.19
codefuse-ai/F2LLM-v2-0.6B2026.06	78.39	71.52	67	72.57	13.61	87.73	65.14
intfloat/e5-mistral-7b-instruct2026.06	78.32	76.32	66.9	61.62	9.51	90.26	63.82
nvidia/NV-Embed-v22026.06	73.64	77.1	66.64	63.02	11.94	90.61	63.82
nvidia/NV-Embed-v12026.06	73.47	75.67	65.69	66.3	14.69	89.94	64.29