PaLM 2 (Unicorn): benchmark results
Google's largest PaLM 2 size tier (above Gecko, Otter, and Bison), announced at I/O in May 2023 and served as Vertex AI's text-unicorn API from November 2023. Provider: Google. Released 2023-11-30. Access: API.
Aggregate Rating 1546 ± 22, rank #1750 of 10916 rated models, from 89 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| HELM Lite - WMT 2014 (de-en) | 23.58 | BLEU-4 (%) | 100 |
| HELM Lite - WMT 2014 (cs-en) | 27.12 | BLEU-4 (%) | 98.9 |
| HELM Lite - WMT 2014 (fr-en) | 27.36 | BLEU-4 (%) | 98.9 |
| HELM Lite - WMT 2014 (hi-en) | 23.77 | BLEU-4 (%) | 98.9 |
| HELM WMT 2014 | 25.95 | BLEU-4 (%) | 98.9 |
| HELM MMLU - Human Sexuality | 92.37 | EM | 94.9 |
| HELM MMLU - Us Foreign Policy | 96 | EM | 93.6 |
| HELM Lite - WMT 2014 (ru-en) | 27.93 | BLEU-4 (%) | 93.3 |
| HELM Lite - LegalBench - Function of Decision Section | 45.23 | Exact Match (%) | 92.2 |
| HELM MMLU - College Biology | 95.14 | EM | 91.7 |
| HELM v2 Lite - LegalBench | 70.25 | Exact Match (%) | 90 |
| HELM MMLU - Business Ethics | 83 | EM | 87.8 |
Interactive version: theaggregate.ai/model?slug=palm-2-unicorn · How It Works · Data refreshed daily, snapshot 2026-11-08.