rembert: benchmark results
Provider: Google. Released 2020-12-01. Access: Open.
Aggregate Rating 1506 ± 37, rank #2259 of 10916 rated models, from 133 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| EuroEval Spanish NLU - CoNLL ES | 84.58 | Micro F1 excluding MISC (%) | 99.5 |
| EuroEval Italian NLU - MultiNERD IT | 88.34 | Micro F1 excluding MISC (%) | 99.3 |
| EuroEval Romanian NLU - RONEC | 76.87 | Micro F1 excluding MISC (%) | 99 |
| EuroEval Estonian NLU - EstNER | 58.76 | Micro F1 excluding MISC (%) | 98.8 |
| EuroEval Greek NLU - ELNER | 81.62 | Micro F1 excluding MISC (%) | 98 |
| EuroEval Czech NLU - PONER | 67.86 | Micro F1 excluding MISC (%) | 97.8 |
| EuroEval Hungarian NLU - Szeged NER | 90.48 | Micro F1 excluding MISC (%) | 97.5 |
| EuroEval Polish NLU - Polemo2 | 96.37 | Matthews correlation coefficient (%) | 97 |
| EuroEval Croatian NLU - WikiANN HR | 73.18 | Micro F1 excluding MISC (%) | 96.7 |
| EuroEval Lithuanian NLU - WikiANN LT | 70.72 | Micro F1 excluding MISC (%) | 96.5 |
| EuroEval Portuguese NLU - HAREM | 64.49 | Micro F1 excluding MISC (%) | 96.3 |
| EuroEval Serbian NLU - UNER SR | 88.69 | Micro F1 excluding MISC (%) | 96.1 |
Interactive version: theaggregate.ai/model?slug=rembert · How It Works · Data refreshed daily, snapshot 2026-11-08.