GPT-4o (2024-11-20): benchmark results
November 20, 2024 GPT-4o snapshot, tracked when sources report the dated API model. Provider: OpenAI. Released 2024-11-20. Access: API.
Aggregate Rating 1622 ± 5, rank #381 of 1656 rated models, from 898 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| ACEBench | 89.6 | Overall (imported) | 100 |
| AgentRewardBench (Accessibility Tree) | 69.8 | Precision (%) | 100 |
| AgentRewardBench (Accessibility Tree) - VisualWebArena | 63 | Precision (%) | 100 |
| AgentRewardBench (Accessibility Tree) - WebArena | 70.2 | Precision (%) | 100 |
| AgentRewardBench (Accessibility Tree) - WorkArena | 94.6 | Precision (%) | 100 |
| AgentRewardBench (Screenshot) - WebArena | 69.9 | Precision (%) | 100 |
| Cedar - Text-Only - Japanese | 42.97 | Accuracy (%) | 100 |
| Cedar - Text-Only - Swahili | 50.43 | Accuracy (%) | 100 |
| ComplexFuncBench | 66.5 | Score (imported) | 100 |
| HELM Robo-Reward-Bench - Robo Reward Bench | 65 | EM | 100 |
| HELM SeaHELM - Flores (en-id) | 70.52 | ChrF++ | 100 |
| HELM SeaHELM - Flores (en-th) | 49.14 | ChrF++ | 100 |
Interactive version: theaggregate.ai/model?slug=gpt-4o-2024-11-20 · How It Works · Data refreshed daily, snapshot 2026-10-20.