ERNIE 4.5: benchmark results
Provider: Baidu. Released 2025-03-16. Access: API.
Unified ELO 1400 ± 1, rank #1134 of 1392 rated models, from 12 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AI for Education Visual Maths - Statistics and Probability | 28.57 | Accuracy (%) | 60.8 |
| ZeroEval GPQA Diamond | 74 | GPQA Diamond Score | 49.8 |
| AI for Education Visual Maths - Geometry | 33.85 | Accuracy (%) | 20.9 |
| AI for Education Visual Maths - Number and Operations | 35.14 | Accuracy (%) | 18.2 |
| AI for Education Visual Maths | 31.65 | Accuracy (%) | 14.9 |
| AI for Education Visual Maths - Algebra | 30.77 | Accuracy (%) | 9.5 |
| AI for Education Visual Maths - Measurement | 21.62 | Accuracy (%) | 6.8 |
| LLM Stats (C-Eval) | 40.7 | Score (%) | 0 |
| LLM Stats (CMMLU) | 39.8 | Score (%) | 0 |
| LLM Stats (DROP) | 28.6 | Score (%) | 0 |
| LLM Stats (MMLU-Redux) | 43.2 | Score (%) | 0 |
| LLM Stats Score | -13.27 | LLM Stats Score (conservative rating) | 0 |
Interactive version: theaggregate.ai/model?slug=ernie-4-5 · How It Works · Data refreshed daily, snapshot 2026-09-05.