PaLM 540B: benchmark results
Provider: Google. Released 2022-04-04. Access: API.
Aggregate Rating 1448 ± 43, rank #3919 of 10552 rated models, from 10 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| BoolQ | 88.7 | Accuracy (%) | 90.3 |
| WinoGrande | 85.1 | Accuracy (%) | 90 |
| ARC Challenge (AI2) | 85.2 | Accuracy (%) | 84.9 |
| Epoch AI - Lambada | 77.9 | Score | 80 |
| HellaSwag | 83.8 | Accuracy (%) | 78.9 |
| OpenBookQA | 68 | Accuracy (%) | 76.2 |
| TriviaQA | 81.4 | Accuracy (%) | 65.8 |
| PIQA | 82.3 | Accuracy (%) | 65 |
| GSM8K | 56.5 | Accuracy (%) | 59.1 |
| MMLU | 69.3 | Accuracy (%) | 54.5 |
Interactive version: theaggregate.ai/model?slug=palm-540b · How It Works · Data refreshed daily, snapshot 2026-11-16.