Benchmark
MTEB(fra, v1) - STS
Embedding benchmarks measure how well models find related items, for example when matching a search to relevant results. Scores are percentages.
- Models measured
- 19
- Scores recorded
- 19
- Sources
- 1
- IQ weight
- 0.95
Score distribution
- 75.7%
- Mean
- 73.8%
- Median
- 5.4%
- Std dev
- 70.2%
- 10th percentile
- 83%
- 90th percentile
- 68.1%
- Min
This benchmark has a weight of 0.95 for embedding models, measured over 19 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.
Leaderboard
| # | Model | Score |
|---|---|---|
| 1 | Qwen3 Embedding 8BSTS· MTEB Leaderboard data | 86.8% |
| 2 | Qwen3 Embedding 4BSTS· MTEB Leaderboard data | 86.2% |
| 3 | Multilingual E5 Large InstructSTS· MTEB Leaderboard data | 83% |
| 4 | E5 Mistral 7B InstructSTS· MTEB Leaderboard data | 82.5% |
| 5 | Paraphrase Multilingual MPNet v2STS· MTEB Leaderboard data | 78.4% |
| 6 | snowflake-arctic-embed-l-v2.0STS· MTEB Leaderboard data | 78% |
| 7 | Granite Embedding 278m MultilingualSTS· MTEB Leaderboard data | 76.5% |
| 8 | E5 Large V2STS· MTEB Leaderboard data | 75.3% |
| 9 | Gte LargeSTS· MTEB Leaderboard data | 75.2% |
| 10 | UAE-Large-V1STS· MTEB Leaderboard data | 73.8% |
| 11 | BGE Large En V1.5STS· MTEB Leaderboard data | 73.8% |
| 12 | Gte BaseSTS· MTEB Leaderboard data | 73.4% |
| 13 | E5 Base V2STS· MTEB Leaderboard data | 73.2% |
| 14 | nomic-embed-text-v1.5STS· MTEB Leaderboard data | 71.8% |
| 15 | nomic-embed-text-v1STS· MTEB Leaderboard data | 71.5% |
| 16 | Qwen3 Embedding 0.6BSTS· MTEB Leaderboard data | 71% |
| 17 | BGE Base En V1.5STS· MTEB Leaderboard data | 70.2% |
| 18 | All MiniLM L6 V2STS· MTEB Leaderboard data | 69.4% |
| 19 | bge-small-en-v1.5STS· MTEB Leaderboard data | 68.1% |