ERNIE-4.0-Turbo-8K-Preview: benchmark results
Provider: Baidu. Released 2024-07-04. Access: API.
Aggregate Rating 1565 ± 22, rank #1385 of 10870 rated models, from 53 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| OpenCompass Reasoning - Humanities - Chinese (CompassBench 2409) | 62.9 | Score (%) | 96.6 |
| OpenCompass Language - Dialogue - Chinese (CompassBench 2409) | 61.3 | Score (%) | 89.7 |
| OpenCompass Math - College - Chinese (CompassBench 2409) | 65.5 | Score (%) | 86.2 |
| OpenCompass Reasoning - Social - Chinese (CompassBench 2409) | 50.3 | Score (%) | 86.2 |
| OpenCompass Code - HumanEval - Chinese (CompassBench 2409) | 80.5 | Score (%) | 81 |
| OpenCompass Reasoning - Common Sense - Chinese (CompassBench 2409) | 51.8 | Score (%) | 79.3 |
| OpenCompass Reasoning - Science and Engineering - English (CompassBench 2409) | 50 | Score (%) | 79.3 |
| OpenCompass Code - Code QA - English (CompassBench 2409) | 58.7 | Score (%) | 75.9 |
| OpenCompass Code - LCBench - Chinese (CompassBench 2409) | 56.4 | Score (%) | 72.4 |
| OpenCompass Code - LCBench - English (CompassBench 2409) | 57.8 | Score (%) | 72.4 |
| OpenCompass LLM - Code (CompassBench 2409) | 63.2 | Score (%) | 72.4 |
| OpenCompass LLM - Math - Chinese (CompassBench 2409) | 72.1 | Score (%) | 72.4 |
Interactive version: theaggregate.ai/model?slug=ernie-4-0-turbo-8k-preview · How It Works · Data refreshed daily, snapshot 2026-11-12.