ERNIE 4.5 300B A47B: benchmark results
Baidu's open ERNIE 4.5 MoE model (300B total, 47B active). Provider: Baidu. Released 2025-03-16. Access: Open.
Aggregate Rating 1596 ± 13, rank #458 of 1656 rated models, from 175 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AGC-Bench - scimon | 4.03 | Dataset z-score | 100 |
| AGC-Bench - chinese_homophonic_puns | 1.38 | Dataset z-score | 97.5 |
| AGC-Bench - simile_generation | 1.62 | Dataset z-score | 96.2 |
| AGC-Bench - fann_or_flop | 1.36 | Dataset z-score | 95.1 |
| AgentDrive - Scenario | 97.5 | Accuracy (%) | 93 |
| AGC-Bench - grapheval_ai_researcher | 1.02 | Dataset z-score | 92.6 |
| AGC-Bench - munch | 0.98 | Dataset z-score | 90.9 |
| AgentDrive - Policy | 95 | Accuracy (%) | 87 |
| AGC-Bench - Figurative Language | 0.71 | JRT z-score | 86.6 |
| AGC-Bench - STEM | 0.53 | JRT z-score | 86.6 |
| AGC-Bench - fig_qa | 1.89 | Dataset z-score | 86.6 |
| AGC-Bench - moh_x | 0.76 | Dataset z-score | 84 |
Interactive version: theaggregate.ai/model?slug=ernie-4-5-300b-a47b · How It Works · Data refreshed daily, snapshot 2026-10-20.