semantic similarity benchmark
MTEB(eng, v2) Semantic Similarity Leaderboard
Direct model-level leaderboard results from MTEB.
Updated Aug 18, 2026
Models100
Model coverage207
MetricOfficial leaderboard task-type mean
EvidenceA
MTEB(eng, v2) Semantic Similarity Ranking
Higher official leaderboard task-type mean ranks better on this benchmark.
Sort by
Rank | Model | Official leaderboard task-type mean | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelKLQZhou-EmbeddingKingsoft Llm | Official leaderboard task-type mean91.6% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank02 | ModelJCingot-8b-r3Jcorners | Official leaderboard task-type mean89.3% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank03 | ModelINJasper-Token-Compression-600MInfgrad | Official leaderboard task-type mean88.8% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank04 | ModelAC | Official leaderboard task-type mean88.7% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank05 | ModelAC | Official leaderboard task-type mean88.6% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank06 | ModelJIjina-embeddings-v5-omni-nanoJinaai | Official leaderboard task-type mean88.3% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank07 | ModelJIjina-embeddings-v5-text-nanoJinaai | Official leaderboard task-type mean88.3% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank08 | ModelJIjina-embeddings-v5-omni-smallJinaai | Official leaderboard task-type mean88.1% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank09 | ModelJIjina-embeddings-v5-text-smallJinaai | Official leaderboard task-type mean88.1% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank10 | ModelSESeed1.5-Embedding | Official leaderboard task-type mean87.2% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank11 | ModelBY | Official leaderboard task-type mean86.9% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank12 | ModelANLGAI-Embedding-PreviewAnnamodels | Official leaderboard task-type mean86.7% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank13 | ModelAC | Official leaderboard task-type mean86.6% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank14 | ModelLAbilingual-embedding-largeLajavaness | Official leaderboard task-type mean86.0% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank15 | ModelJIjina-embeddings-v4Jinaai | Official leaderboard task-type mean85.9% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank16 | ModelJIjina-embeddings-v3Jinaai | Official leaderboard task-type mean85.8% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank17 | ModelKINano-Em1-0.6B-v2.1Kitefishai | Official leaderboard task-type mean85.4% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank18 | ModelGO | Official leaderboard task-type mean85.3% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank19 | ModelGO | Official leaderboard task-type mean85.2% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank20 | ModelIEYuan-embedding-2.0-enIeityuan | Official leaderboard task-type mean84.9% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank21 | ModelGO | Official leaderboard task-type mean84.8% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank22 | ModelKEKaLM-embedding-multilingual-mini-instruct-v2.5Kalm Embedding | Official leaderboard task-type mean84.8% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank23 | ModelSASFR-Embedding-MistralSalesforce | Official leaderboard task-type mean84.8% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank24 | ModelHCspeed-embedding-7b-instructHaon Chen | Official leaderboard task-type mean84.7% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank25 | ModelINmultilingual-e5-large-instructIntfloat | Official leaderboard task-type mean84.7% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank26 | ModelLALinq-Embed-MistralLinq Ai Research | Official leaderboard task-type mean84.7% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank27 | ModelCAF2LLM-v2-8BCodefuse Ai | Official leaderboard task-type mean84.6% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank28 | ModelBIBidirLM-Omni-2.5B-EmbeddingBidirlm | Official leaderboard task-type mean84.6% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank29 | ModelTATarka-Embedding-350M-V1Tarka Air | Official leaderboard task-type mean84.6% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank30 | ModelCAF2LLM-v2-14BCodefuse Ai | Official leaderboard task-type mean84.5% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank31 | ModelAVGIST-large-Embedding-v0Avsolatorio | Official leaderboard task-type mean84.4% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank32 | ModelCAF2LLM-v2-1.7BCodefuse Ai | Official leaderboard task-type mean84.4% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank33 | ModelMAmxbai-embed-large-v1Mixedbread Ai | Official leaderboard task-type mean84.4% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank34 | ModelNOjasper_en_vision_language_v1Novasearch | Official leaderboard task-type mean84.4% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank35 | ModelWHUAE-Large-V1Whereisai | Official leaderboard task-type mean84.4% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank36 | ModelINe5-mistral-7b-instructIntfloat | Official leaderboard task-type mean84.3% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank37 | ModelCAF2LLM-v2-4BCodefuse Ai | Official leaderboard task-type mean84.2% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank38 | ModelCAF2LLM-4BCodefuse Ai | Official leaderboard task-type mean84.2% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank39 | ModelVOvoyage-3-m-expVoyageai | Official leaderboard task-type mean84.0% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank40 | ModelGO | Official leaderboard task-type mean83.9% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank41 | ModelNOstella_en_400M_v5Novasearch | Official leaderboard task-type mean83.9% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank42 | ModelCAF2LLM-1.7BCodefuse Ai | Official leaderboard task-type mean83.9% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank43 | ModelNV | Official leaderboard task-type mean83.8% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank44 | ModelMNLLM2Vec-Mistral-7B-Instruct-v2-mntp-supervisedMcgill Nlp | Official leaderboard task-type mean83.7% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank45 | ModelGO | Official leaderboard task-type mean83.6% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank46 | ModelCAF2LLM-v2-0.6BCodefuse Ai | Official leaderboard task-type mean83.5% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank47 | ModelLELCO-Embedding-Omni-7BLco Embedding | Official leaderboard task-type mean83.4% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank48 | ModelAVGIST-Embedding-v0Avsolatorio | Official leaderboard task-type mean83.3% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank49 | ModelNOstella_en_1.5B_v5Novasearch | Official leaderboard task-type mean83.3% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank50 | ModelTHgte-largeThenlper | Official leaderboard task-type mean83.3% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank51 | ModelMAdeepset-mxbai-embed-de-large-v1Mixedbread Ai | Official leaderboard task-type mean83.1% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank52 | ModelCO | Official leaderboard task-type mean83.0% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank53 | ModelGRGritLM-7BGritlm | Official leaderboard task-type mean83.0% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank54 | ModelGRGritLM-8x7BGritlm | Official leaderboard task-type mean82.9% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank55 | ModelASGiga-Embeddings-instructAi Sage | Official leaderboard task-type mean82.9% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank56 | ModelLELCO-Embedding-Omni-3BLco Embedding | Official leaderboard task-type mean82.8% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank57 | ModelCAF2LLM-v2-330MCodefuse Ai | Official leaderboard task-type mean82.8% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank58 | ModelMAminiac-embedManiaclabs | Official leaderboard task-type mean82.8% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank59 | ModelVOvoyage-large-2-instructVoyageai | Official leaderboard task-type mean82.8% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank60 | ModelAVGIST-small-Embedding-v0Avsolatorio | Official leaderboard task-type mean82.8% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank61 | ModelLAbilingual-embedding-baseLajavaness | Official leaderboard task-type mean82.8% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank62 | ModelBAbge-large-en-v1.5Baai | Official leaderboard task-type mean82.8% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank63 | ModelAVNoInstruct-small-Embedding-v0Avsolatorio | Official leaderboard task-type mean82.8% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank64 | ModelINstella-base-en-v2Infgrad | Official leaderboard task-type mean82.7% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank65 | ModelANgte-Qwen2-7B-instructAlibaba Nlp | Official leaderboard task-type mean82.7% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank66 | ModelMOmdbr-leaf-mtMongodb | Official leaderboard task-type mean82.6% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank67 | ModelANgte-Qwen2-1.5B-instructAlibaba Nlp | Official leaderboard task-type mean82.5% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank68 | ModelCAF2LLM-0.6BCodefuse Ai | Official leaderboard task-type mean82.5% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank69 | ModelCO | Official leaderboard task-type mean82.4% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank70 | ModelDEUSER-bge-m3Deepvk | Official leaderboard task-type mean82.4% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank71 | ModelVOvoyage-code-3Voyageai | Official leaderboard task-type mean82.4% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank72 | ModelITBOOM_4B_v1Ict Time And Querit | Official leaderboard task-type mean82.3% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank73 | ModelNV | Official leaderboard task-type mean82.2% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank74 | ModelTHgte-baseThenlper | Official leaderboard task-type mean82.2% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank75 | ModelHTKaLM-embedding-multilingual-mini-v1Hit Tmg | Official leaderboard task-type mean82.2% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank76 | ModelBAbge-base-en-v1.5Baai | Official leaderboard task-type mean82.1% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank77 | ModelANgte-Qwen1.5-7B-instructAlibaba Nlp | Official leaderboard task-type mean82.1% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank78 | ModelINe5-largeIntfloat | Official leaderboard task-type mean82.0% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank79 | ModelHTKaLM-embedding-multilingual-mini-instruct-v1Hit Tmg | Official leaderboard task-type mean82.0% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank80 | ModelSDmmlw-e5-baseSdadas | Official leaderboard task-type mean81.9% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank81 | ModelTHgte-smallThenlper | Official leaderboard task-type mean81.9% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank82 | ModelANgte-multilingual-baseAlibaba Nlp | Official leaderboard task-type mean81.9% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank83 | ModelSDmmlw-roberta-baseSdadas | Official leaderboard task-type mean81.8% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank84 | ModelNAnomic-embed-text-v1Nomic Ai | Official leaderboard task-type mean81.8% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank85 | ModelITICT-TIME-and-Querit-embedding-v1Ict Time And Querit | Official leaderboard task-type mean81.8% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank86 | ModelVPSearchMap_PreviewVplabs | Official leaderboard task-type mean81.6% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank87 | ModelPRmini-gtePrdev | Official leaderboard task-type mean81.6% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank88 | ModelTATarka-Embedding-150M-V1Tarka Air | Official leaderboard task-type mean81.5% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank89 | ModelLAbilingual-embedding-smallLajavaness | Official leaderboard task-type mean81.5% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank90 | ModelOP | Official leaderboard task-type mean81.4% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank91 | ModelKINano-Em1-0.6B-v2Kitefishai | Official leaderboard task-type mean81.4% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank92 | ModelINmultilingual-e5-largeIntfloat | Official leaderboard task-type mean81.4% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank93 | ModelOP | Official leaderboard task-type mean81.4% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank94 | ModelSDmmlw-e5-largeSdadas | Official leaderboard task-type mean81.4% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank95 | ModelABMedEmbed-small-v0.1Abhinand | Official leaderboard task-type mean81.3% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank96 | ModelBAbge-small-en-v1.5Baai | Official leaderboard task-type mean81.3% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank97 | ModelSDmmlw-roberta-largeSdadas | Official leaderboard task-type mean81.2% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank98 | ModelMAbge-m3-custom-frManu | Official leaderboard task-type mean81.2% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank99 | ModelSASFR-Embedding-2_RSalesforce | Official leaderboard task-type mean80.9% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
| Rank100 | ModelCO | Official leaderboard task-type mean80.8% | PercentileN/A | Participants207 | EvidenceA | Evaluated |
MTEB(eng, v2) Semantic Similarity Highlights
The leading models and scores on this benchmark.
MTEB(eng, v2) Semantic Similarity Score Distribution
A closer view of the leading scores on this benchmark.
The Top AI Models for MTEB(eng, v2) Semantic Similarity
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
Ranking basisThis mteb(eng, v2) semantic similarity AI model leaderboard uses descending official leaderboard task-type mean in the benchmark's original unit. The leaderboard ranking keeps matched price and speed data separate from benchmark evidence.
- 01KLQZhou-EmbeddingKingsoft Llm
- Official leaderboard task-type mean
- 91.6%
Strengths
- Ranks #1 of 207 compared models
- A evidence result
Considerations
- This result measures MTEB(eng, v2) Semantic Similarity, not total model capability
- 02JCingot-8b-r3Jcorners
- Official leaderboard task-type mean
- 89.3%
Strengths
- Ranks #2 of 207 compared models
- A evidence result
Considerations
- This result measures MTEB(eng, v2) Semantic Similarity, not total model capability
- 03IN
- Official leaderboard task-type mean
- 88.8%
Strengths
- Ranks #3 of 207 compared models
- A evidence result
Considerations
- This result measures MTEB(eng, v2) Semantic Similarity, not total model capability
- 04ACQwen3-Embedding-4BAlibaba Cloud / Qwen Team
- Official leaderboard task-type mean
- 88.7%
Strengths
- Ranks #4 of 207 compared models
- A evidence result
Considerations
- This result measures MTEB(eng, v2) Semantic Similarity, not total model capability
- 05ACQwen3-Embedding-8BAlibaba Cloud / Qwen Team
- Official leaderboard task-type mean
- 88.6%
Strengths
- Ranks #5 of 207 compared models
- A evidence result
Considerations
- This result measures MTEB(eng, v2) Semantic Similarity, not total model capability
Selection summary
Best AI Models for MTEB(eng, v2) Semantic Similarity
QZhou-Embedding currently leads MTEB(eng, v2) Semantic Similarity with 91.6%. 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.
What is MTEB(eng, v2) Semantic Similarity?
What MTEB(eng, v2) Semantic Similarity measures and how its scores work.
Direct model-level leaderboard results from MTEB.
Scores are shown in ratio. This benchmark is verified and has an evidence level of A.
- Family
- MTEB(eng, v2):semantic_similarity
- Modality
- multimodal
- Primary category
- semantic similarity
- Score direction
- higher
- LLMBoard eligible
- Yes
- Evaluation key
- mteb-mteb-eng-v2-semantic_similarity|mteb|current
Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about MTEB(eng, v2) Semantic Similarity.
Which model scores highest on MTEB(eng, v2) Semantic Similarity?
QZhou-Embedding is currently ranked first with 91.6%.
What does MTEB(eng, v2) Semantic Similarity measure?
Direct model-level leaderboard results from MTEB.
Is a higher official leaderboard task-type mean better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
100 model results are currently shown.
Does this benchmark affect the overall score?
Yes. This benchmark can contribute to the current LLMBoard capability score.