SkMTEB - Retrieval: leaderboard
SkMTEB is the Slovak Massive Text Embedding Benchmark: 31 datasets across seven task types (bitext mining, classification, clustering, pair classification, reranking, retrieval, semantic textual similarity). This board reports Mean score (%) over the five retrieval datasets; higher is better.
Metric: Mean score (%) over the five retrieval datasets; higher is better. Source: arxiv.org. Saturation forecast: Rough model projection: around 2026. 28 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | gemini-embedding-001 | 88.34 | |
| 2 | Qwen3-Embedding-8B | 86.27 | |
| 3 | multilingual-e5-large-instruct | 86.08 | |
| 4 | multilingual-e5-large | 85.8 | |
| 5 | jina-embeddings-v3 | 85.77 | |
| 6 | embed-v4.0 | 85.71 | |
| 7 | bge-m3 | 85.63 | |
| 8 | text-embedding-3-large | 85.55 | |
| 9 | nomic-embed-text-v2-moe | 85.37 | |
| 10 | snowflake-arctic-embed-l-v2.0 | 85.17 |
Interactive version: theaggregate.ai/benchmark?slug=skmteb-retrieval · How It Works · Data refreshed daily, snapshot 2026-11-13.