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 2023-05-10. Access: API.

Unified ELO 1412 ± 72, rank #1170 of 1776 rated models, from 9 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
HELM WMT 201425.95BLEU-4 (%)98.9
HELM v2 Lite - LegalBench70.25Exact Match (%)90
HELM v2 Lite - MedQA76.92Exact Match (%)87.5
HELM NaturalQuestions (Closed)43.48F1 (%)78.9
HELM Lite68.11Mean win rate (self-reported)74
HELM (Stanford)64.36Mean Win Rate (%)68.9
HELM NaturalQuestions (Open)67.42F1 (%)31.1
HELM v2 Lite - HumanEval (Code)0Pass@1 (%)11.8
HELM NarrativeQA58.29F1 (%)11.1

Interactive version: theaggregate.ai/model?slug=palm-2-unicorn · How the rankings work · Data refreshed daily, snapshot 2026-07-22.