ERNIE 5.0 Thinking Preview: benchmark results
Baidu's extended-thinking preview variant of the multimodal ERNIE 5.0 flagship. Provider: Baidu. Released 2026-01-01. Access: API.
Aggregate Rating 1670 ± 26, rank #268 of 1656 rated models, from 74 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA LiveCodeBench | 81.16 | Pass@1 (%) | 93.3 |
| AGI-Eval Community - General Reasoning | 97.41 | Accuracy (%) | 90.7 |
| AA MMLU-Pro | 83.03 | Accuracy (%) | 83 |
| AA AIME 2025 | 85 | Accuracy (%) | 82.8 |
| ReLE - Law and Civil Service | 82.3 | Accuracy (%) | 76.1 |
| Tau2-Bench Telecom | 83.92 | Success Rate (%) | 75.1 |
| AA TAU-2 Bench | 83.92 | Accuracy (%) | 74.8 |
| AGI-Eval Community - Mathematical Reasoning | 79.93 | Accuracy (%) | 70.7 |
| ReLE - Law - Lawyer Qualification Exam | 74.7 | Accuracy (%) | 66.6 |
| AA Terminal-Bench Hard | 25 | Accuracy (%) | 66 |
| AGI-Eval Community - Subject Knowledge | 85.15 | Accuracy (%) | 65.7 |
| AGI-Eval Community - Subject Reasoning (Chinese) | 86.36 | Accuracy (%) | 65.6 |
Interactive version: theaggregate.ai/model?slug=ernie-5-0-thinking-preview · How It Works · Data refreshed daily, snapshot 2026-10-20.