semantic similarity benchmark
MIEB English Semantic Similarity is a current leaderboard for MIEB(eng) semantic_similarity.
Updated Sep 8, 2026
Higher official leaderboard task-type mean ranks better on this benchmark.
Rank | Model | Official leaderboard task-type mean | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelLALanguageBind_OmniLanguagebind | Official leaderboard task-type mean69.58% | Percentile100.00% | Participants50 | EvidenceA | EvaluatedN/A |
| Rank02 | ModelJIjina-embeddings-v5-omni-smallJinaai | Official leaderboard task-type mean63.93% | Percentile97.96% | Participants50 | EvidenceA | EvaluatedN/A |
| Rank03 | ModelDNDNA-VL-STEER-2BDnotitia | Official leaderboard task-type mean62.84% | Percentile95.92% | Participants50 | EvidenceA | EvaluatedN/A |
| Rank04 | ModelROe5-vRoyokong | Official leaderboard task-type mean62.80% | Percentile93.88% | Participants50 | EvidenceA | EvaluatedN/A |
| Rank05 | ModelVOvoyage-multimodal-3Voyageai | Official leaderboard task-type mean62.68% | Percentile91.84% | Participants50 | EvidenceA | EvaluatedN/A |
| Rank06 | ModelQUEVA02-CLIP-bigE-14-plusQuansun | Official leaderboard task-type mean58.83% | Percentile89.80% | Participants50 | EvidenceA | EvaluatedN/A |
| Rank07 | ModelLACLIP-ViT-bigG-14-laion2B-39B-b160kLaion | Official leaderboard task-type mean56.64% | Percentile87.76% | Participants50 | EvidenceA | EvaluatedN/A |
| Rank08 | ModelOP | Official leaderboard task-type mean55.61% | Percentile85.71% | Participants50 | EvidenceA | EvaluatedN/A |
| Rank09 | ModelQUEVA02-CLIP-bigE-14Quansun | Official leaderboard task-type mean55.57% | Percentile83.67% | Participants50 | EvidenceA | EvaluatedN/A |
| Rank10 | ModelLACLIP-ViT-g-14-laion2B-s34B-b88KLaion | Official leaderboard task-type mean55.54% | Percentile81.63% | Participants50 | EvidenceA | EvaluatedN/A |
| Rank11 | ModelJIjina-embeddings-v5-omni-nanoJinaai | Official leaderboard task-type mean54.76% | Percentile79.59% | Participants50 | EvidenceA | EvaluatedN/A |
| Rank12 | ModelOP | Official leaderboard task-type mean54.60% | Percentile77.55% | Participants50 | EvidenceA | EvaluatedN/A |
| Rank13 | ModelGO | Official leaderboard task-type mean54.60% | Percentile75.51% | Participants50 | EvidenceA | EvaluatedN/A |
| Rank14 | ModelLACLIP-ViT-L-14-DataComp.XL-s13B-b90KLaion | Official leaderboard task-type mean54.49% | Percentile73.47% | Participants50 | EvidenceA | EvaluatedN/A |
| Rank15 | ModelSAblip-itm-base-cocoSalesforce | Official leaderboard task-type mean54.28% | Percentile71.43% | Participants50 | EvidenceA | EvaluatedN/A |
| Rank16 | ModelGO | Official leaderboard task-type mean54.19% | Percentile69.39% | Participants50 | EvidenceA | EvaluatedN/A |
| Rank17 | ModelGO | Official leaderboard task-type mean54.09% | Percentile67.35% | Participants50 | EvidenceA | EvaluatedN/A |
| Rank18 | ModelTLVLM2Vec-FullTiger Lab | Official leaderboard task-type mean54.04% | Percentile65.31% | Participants50 | EvidenceA | EvaluatedN/A |
| Rank19 | ModelJIjina-clip-v1Jinaai | Official leaderboard task-type mean53.86% | Percentile63.27% | Participants50 | EvidenceA | EvaluatedN/A |
| Rank20 | ModelLACLIP-ViT-H-14-laion2B-s32B-b79KLaion | Official leaderboard task-type mean53.77% | Percentile61.22% | Participants50 | EvidenceA | EvaluatedN/A |
| Rank21 | ModelTLVLM2Vec-LoRATiger Lab | Official leaderboard task-type mean53.60% | Percentile59.18% | Participants50 | EvidenceA | EvaluatedN/A |
| Rank22 | ModelQUEVA02-CLIP-L-14Quansun | Official leaderboard task-type mean53.59% | Percentile57.14% | Participants50 | EvidenceA | EvaluatedN/A |
| Rank23 | ModelLACLIP-ViT-B-16-DataComp.XL-s13B-b90KLaion | Official leaderboard task-type mean53.56% | Percentile55.10% | Participants50 | EvidenceA | EvaluatedN/A |
| Rank24 | ModelSAblip-itm-base-flickrSalesforce | Official leaderboard task-type mean53.49% | Percentile53.06% | Participants50 | EvidenceA | EvaluatedN/A |
| Rank25 | ModelLACLIP-ViT-L-14-laion2B-s32B-b82KLaion | Official leaderboard task-type mean53.33% | Percentile51.02% | Participants50 | EvidenceA | EvaluatedN/A |
| Rank26 | ModelGO | Official leaderboard task-type mean52.84% | Percentile48.98% | Participants50 | EvidenceA | EvaluatedN/A |
| Rank27 | ModelGO | Official leaderboard task-type mean52.80% | Percentile46.94% | Participants50 | EvidenceA | EvaluatedN/A |
| Rank28 | ModelGO | Official leaderboard task-type mean52.60% | Percentile44.90% | Participants50 | EvidenceA | EvaluatedN/A |
| Rank29 | ModelGO | Official leaderboard task-type mean52.04% | Percentile42.86% | Participants50 | EvidenceA | EvaluatedN/A |
| Rank30 | ModelGO | Official leaderboard task-type mean51.80% | Percentile40.82% | Participants50 | EvidenceA | EvaluatedN/A |
The leading models and scores on this benchmark.
A closer view of the leading scores on this benchmark.
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
What MIEB English Semantic Similarity measures and how its scores work.
A benchmark leaderboard for MIEB(eng) semantic_similarity.
The Official leaderboard task-type mean, reported as a ratio.
Scores are shown in ratio. This benchmark is verified and has an evidence level of A.
MTEB. Benchmark scores retain their original unit. Overall score eligibility is shown separately.
Common questions about MIEB English Semantic Similarity.
LanguageBind_Omni is currently ranked first with 69.58%.
The current leaders are LanguageBind_Omni (69.58%), jina-embeddings-v5-omni-small (63.93%), and DNA-VL-STEER-2B (62.84%).
No matched official input price is currently available.
No matched runtime record is currently available.
No. This benchmark measures one defined capability or task. The overall LLMBoard score uses a separate aggregation across eligible benchmark evidence.
The Official leaderboard task-type mean, reported as a ratio.
Yes. Higher values rank better for this benchmark.
50 model results are currently shown.
No. This benchmark is shown for reference but does not contribute to the overall score.
Ranking basisThis mieb english semantic similarity AI model leaderboard uses descending official leaderboard task-type mean in the benchmark original unit. The leaderboard ranking keeps matched price and speed data separate from benchmark evidence.
Selection summary
LanguageBind_Omni currently leads MIEB English Semantic Similarity with 69.58%. It is the top model on this specific benchmark, while the best LLM for the broader task should also be checked against other benchmarks, price, and runtime.
Use this leaderboard with the supporting benchmark results and coverage details above. A leaderboard position summarizes the selected ranking signal; it does not replace workload-specific testing.