Skip to content

Benchmark

Agentset Embedding Elo - MSMARCO

Embedding benchmarks measure how well models find related items, for example when matching a search to relevant results. Higher scores are better.

Models measured
15
Scores recorded
15
Sources
1
IQ weight
No weight published

Score distribution

1,500
Mean
1,500
Median
0
Std dev
1,500
10th percentile
1,500
90th percentile
1,500
Min

Leaderboard

#ModelScore
1BGE M3MSMARCO· Agentset1,500
2Embed Multilingual V3.0MSMARCO· Agentset1,500
3Gemini Embedding 2MSMARCO· Agentset1,500
4Jina Embeddings v3MSMARCO· Agentset1,500
5Jina Embeddings v5 Text SmallMSMARCO· Agentset1,500
6Qwen3 Embedding 0.6BMSMARCO· Agentset1,500
7Qwen3 Embedding 4BMSMARCO· Agentset1,500
8Qwen3 Embedding 8BMSMARCO· Agentset1,500
9Text Embedding 004MSMARCO· Agentset1,500
10Text Embedding 3 LargeMSMARCO· Agentset1,500
11Text Embedding 3 SmallMSMARCO· Agentset1,500
12voyage-3-largeMSMARCO· Agentset1,500
13voyage-3.5MSMARCO· Agentset1,500
14voyage-3.5-liteMSMARCO· Agentset1,500
15voyage-4MSMARCO· Agentset1,500