multilingual-e5-large: benchmark results
Provider: Microsoft. Released 2022-12-01. Access: Open.
Aggregate Rating 1541 ± 33, rank #1822 of 10916 rated models, from 162 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| EuroEval Albanian NLU - WikiANN SQ | 86.6 | Micro F1 excluding MISC (%) | 100 |
| EuroEval Bosnian NLU - WikiANN BS | 84.87 | Micro F1 excluding MISC (%) | 100 |
| EuroEval Croatian NLU - MMS HR | 49.96 | Matthews correlation coefficient (%) | 100 |
| EuroEval Greek NLU - ELNER | 83.55 | Micro F1 excluding MISC (%) | 100 |
| EuroEval Italian NLU - MultiNERD IT | 90.74 | Micro F1 excluding MISC (%) | 99.8 |
| EuroEval Lithuanian NLU - WikiANN LT | 78.56 | Micro F1 excluding MISC (%) | 99.6 |
| EuroEval Croatian NLU - WikiANN HR | 80.45 | Micro F1 excluding MISC (%) | 99.5 |
| EuroEval Spanish NLU - CoNLL ES | 84.4 | Micro F1 excluding MISC (%) | 99.3 |
| EuroEval Catalan NLU - WikiANN CA | 82.78 | Micro F1 excluding MISC (%) | 99.1 |
| EuroEval Serbian NLU - UNER SR | 90.42 | Micro F1 excluding MISC (%) | 99 |
| EuroEval Portuguese NLU - HAREM | 74.37 | Micro F1 excluding MISC (%) | 98.8 |
| EuroEval Romanian NLU - RONEC | 76.62 | Micro F1 excluding MISC (%) | 98.5 |
Interactive version: theaggregate.ai/model?slug=multilingual-e5-large · How It Works · Data refreshed daily, snapshot 2026-11-08.