clustering benchmark
MTEB English v2 Clustering is the current MTEB English v2 clustering leaderboard.
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 | ModelCAF2LLM-4BCodefuse Ai | Official leaderboard task-type mean68.54% | Percentile100.00% | Participants217 | EvidenceA | EvaluatedN/A |
| Rank02 | ModelCAF2LLM-1.7BCodefuse Ai | Official leaderboard task-type mean64.13% | Percentile99.54% | Participants217 | EvidenceA | EvaluatedN/A |
| Rank03 | ModelCAF2LLM-v2-14BCodefuse Ai | Official leaderboard task-type mean63.54% | Percentile99.07% | Participants217 | EvidenceA | EvaluatedN/A |
| Rank04 | ModelCAF2LLM-v2-8BCodefuse Ai | Official leaderboard task-type mean62.85% | Percentile98.61% | Participants217 | EvidenceA | EvaluatedN/A |
| Rank05 | ModelCAF2LLM-v2-4BCodefuse Ai | Official leaderboard task-type mean62.03% | Percentile98.15% | Participants217 | EvidenceA | EvaluatedN/A |
| Rank06 | ModelKLQZhou-EmbeddingKingsoft Llm | Official leaderboard task-type mean61.65% | Percentile97.69% | Participants217 | EvidenceA | EvaluatedN/A |
| Rank07 | ModelCAF2LLM-v2-1.7BCodefuse Ai | Official leaderboard task-type mean60.91% | Percentile97.22% | Participants217 | EvidenceA | EvaluatedN/A |
| Rank08 | ModelANgte-Qwen1.5-7B-instructAlibaba Nlp | Official leaderboard task-type mean60.91% | Percentile96.76% | Participants217 | EvidenceA | EvaluatedN/A |
| Rank09 | ModelBSSeed1.5-EmbeddingBytedance Seed | Official leaderboard task-type mean60.83% | Percentile96.30% | Participants217 | EvidenceA | EvaluatedN/A |
| Rank10 | ModelNOjasper_en_vision_language_v1Novasearch | Official leaderboard task-type mean60.52% | Percentile95.83% | Participants217 | EvidenceA | EvaluatedN/A |
| Rank11 | ModelCAF2LLM-0.6BCodefuse Ai | Official leaderboard task-type mean60.36% | Percentile95.37% | Participants217 | EvidenceA | EvaluatedN/A |
| Rank12 | ModelINJasper-Token-Compression-600MInfgrad | Official leaderboard task-type mean59.44% | Percentile94.91% | Participants217 | EvidenceA | EvaluatedN/A |
| Rank13 | ModelGO | Official leaderboard task-type mean59.39% | Percentile94.44% | Participants217 | EvidenceA | EvaluatedN/A |
| Rank14 | ModelSASFR-Embedding-2_RSalesforce | Official leaderboard task-type mean59.39% | Percentile93.98% | Participants217 | EvidenceA | EvaluatedN/A |
| Rank15 | ModelANLGAI-Embedding-PreviewAnnamodels | Official leaderboard task-type mean59.25% | Percentile93.52% | Participants217 | EvidenceA | EvaluatedN/A |
| Rank16 | ModelBY | Official leaderboard task-type mean59.22% | Percentile93.06% | Participants217 | EvidenceA | EvaluatedN/A |
| Rank17 | ModelASGiga-Embeddings-instruct-10B-A1.8B-0826Ai Sage | Official leaderboard task-type mean59.07% | Percentile92.59% | Participants217 | EvidenceA | EvaluatedN/A |
| Rank18 | ModelANgte-Qwen2-7B-instructAlibaba Nlp | Official leaderboard task-type mean58.97% | Percentile92.13% | Participants217 | EvidenceA | EvaluatedN/A |
| Rank19 | ModelAC | Official leaderboard task-type mean58.57% | Percentile91.67% | Participants217 | EvidenceA | EvaluatedN/A |
| Rank20 | ModelJCingot-8b-r3Jcorners | Official leaderboard task-type mean58.47% | Percentile91.20% | Participants217 | EvidenceA | EvaluatedN/A |
| Rank21 | ModelCAF2LLM-v2-0.6BCodefuse Ai | Official leaderboard task-type mean58.24% | Percentile90.74% | Participants217 | EvidenceA | EvaluatedN/A |
| Rank22 | ModelKEKaLM-embedding-multilingual-mini-instruct-v2.5Kalm Embedding | Official leaderboard task-type mean58.12% | Percentile90.28% | Participants217 | EvidenceA | EvaluatedN/A |
| Rank23 | ModelASGiga-Embeddings-instruct-3B-0826Ai Sage | Official leaderboard task-type mean57.95% | Percentile89.81% | Participants217 | EvidenceA | EvaluatedN/A |
| Rank24 | ModelNOstella_en_400M_v5Novasearch | Official leaderboard task-type mean57.65% | Percentile89.35% | Participants217 | EvidenceA | EvaluatedN/A |
| Rank25 | ModelAC | Official leaderboard task-type mean57.51% | Percentile88.89% | Participants217 | EvidenceA | EvaluatedN/A |
| Rank26 | ModelNOstella_en_1.5B_v5Novasearch | Official leaderboard task-type mean57.06% | Percentile88.43% | Participants217 | EvidenceA | EvaluatedN/A |
| Rank27 | ModelITBOOM_4B_v1Ict Time And Querit | Official leaderboard task-type mean56.60% | Percentile87.96% | Participants217 | EvidenceA | EvaluatedN/A |
| Rank28 | ModelGO | Official leaderboard task-type mean56.55% | Percentile87.50% | Participants217 | EvidenceA | EvaluatedN/A |
| Rank29 | ModelCAF2LLM-v2-330MCodefuse Ai | Official leaderboard task-type mean56.55% | Percentile87.04% | Participants217 | EvidenceA | EvaluatedN/A |
| Rank30 | ModelGRGeoEmbeddingGeogpt Research Project | Official leaderboard task-type mean56.50% | Percentile86.57% | Participants217 | 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 MTEB English v2 Clustering measures and how its scores work.
A benchmark for the MTEB English v2 clustering task type.
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 MTEB English v2 Clustering.
F2LLM-4B is currently ranked first with 68.54%.
The current leaders are F2LLM-4B (68.54%), F2LLM-1.7B (64.13%), and F2LLM-v2-14B (63.54%).
Gemini Embedding 001 has the lowest matched official input price at $0.15 input per 1M tokens.
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.
100 model results are currently shown.
No. This benchmark is shown for reference but does not contribute to the overall score.
Ranking basisThis mteb english v2 clustering 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
F2LLM-4B currently leads MTEB English v2 Clustering with 68.54%. 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.