PaLM 2 (Bison): benchmark results
Google's mid-size PaLM 2 tier behind Vertex AI's chat-bison API, announced at I/O in May 2023 and since retired. Provider: Google. Released 2023-05-10. Access: API.
Aggregate Rating 1461 ± 22, rank #3288 of 10916 rated models, from 90 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| HELM Lite - LegalBench - Function of Decision Section | 46.59 | Exact Match (%) | 99.4 |
| HELM Lite - WMT 2014 (fr-en) | 25.4 | BLEU-4 (%) | 97.8 |
| HELM Lite - WMT 2014 (hi-en) | 21.98 | BLEU-4 (%) | 97.8 |
| HELM NaturalQuestions (Open) | 81.26 | F1 (%) | 97.8 |
| HELM WMT 2014 | 24.08 | BLEU-4 (%) | 97.8 |
| HELM Lite - WMT 2014 (cs-en) | 25.54 | BLEU-4 (%) | 95.6 |
| HELM Lite - WMT 2014 (de-en) | 22.17 | BLEU-4 (%) | 94.4 |
| HELM v2 Lite - LegalBench | 62.96 | Exact Match (%) | 85 |
| HELM MMLU - World Religions | 88.3 | EM | 79.5 |
| HELM v2 Lite - HumanEval (Code) | 29 | Pass@1 (%) | 76.5 |
| HELM MMLU - Sociology | 92.04 | EM | 76.3 |
| HELM Lite - WMT 2014 (ru-en) | 25.32 | BLEU-4 (%) | 73.3 |
Interactive version: theaggregate.ai/model?slug=palm-2-bison · How It Works · Data refreshed daily, snapshot 2026-11-08.