Benchmark Stats
Models6
Papers6
Metrics1
SOTA History
avg-score
Higher is better
| Rank | Model | Source | Score | Year | Paper |
|---|---|---|---|---|---|
| 1 | NV-Embed-v2 NV-Embed-v2 ranked #1 on MTEB English (56 tasks) as of Aug 2024. Score from official HF model card and paper. | Community | 72.31 | 2024 | Source |
| 2 | GTE-Qwen2-7B-instruct GTE-Qwen2-7B-instruct MTEB English average. Ranked #1 English, #1 Chinese as of Jun 2024. | Community | 72.05 | 2024 | Source |
| 3 | voyage-3-large voyage-3-large MTEB average. Voyage AI blog Jan 2025 reports top ranking; score from MTEB leaderboard snapshot. | Community | 70.32 | 2025 | Source |
| 4 | E5-Mistral-7B-instruct E5-Mistral-7B MTEB average (56 tasks) from the original paper Table 1. | Community | 66.63 | 2024 | Source |
| 5 | jina-embeddings-v3 jina-embeddings-v3 MTEB English average from paper Table 3. | Community | 65.18 | 2024 | Source |
| 6 | text-embedding-3-large OpenAI text-embedding-3-large MTEB score reported in launch blog and confirmed by multiple benchmark comparisons. | Community | 64.6 | 2024 | Source |