OpenAIGPT
GPT-5.4 Pro
OpenAI · GPT. Released Mar 5, 2026.
GPT-5.4 Pro is a model from OpenAI in the GPT family, released Mar 5, 2026. evals.report tracks 13 reported GPT-5.4 Pro benchmark scores across FrontierMath, ARC-AGI-1, ARC-AGI-2, GPQA Diamond, SimpleQA Verified, Epoch Capabilities Index, GAIA: A Benchmark for General AI Assistants, MultiChallenge, and 5 more — each shown with its benchmark, metric, source status, and date, and never combined into a single ranking.
13 results
Benchmark results 13
Compare this model| Benchmark | Category | Score | Metric | Status | Date | |
|---|---|---|---|---|---|---|
| FrontierMath | Reasoning | 50.0% | accuracy | Official | Mar 5, 2026 | Details |
| ARC-AGI-1 | Reasoning | 94.5% | accuracy | Official | Mar 5, 2026 | Details |
| ARC-AGI-2 | Reasoning | 83.33% | accuracy | Official | Mar 5, 2026 | Details |
| GPQA Diamond | Reasoning | 94.6% | accuracy | Official | Mar 5, 2026 | Details |
| SimpleQA Verified | Other | 47.8% | accuracy | Official | Mar 5, 2026 | Details |
| Epoch Capabilities Index | Reasoning | 157.7 | Index | Official | Mar 5, 2026 | Details |
| GAIA: A Benchmark for General AI Assistants | Agents | 50.5% | accuracy | Unverified | Mar 5, 2026 | Details |
| MultiChallenge | Reasoning | 69.23% | accuracy | Verified | Mar 5, 2026 | Details |
| MASK (Model Alignment between Statements and Knowledge) | Other | 91.73 | Honesty score | Verified | Mar 5, 2026 | Details |
| EnigmaEval | Reasoning | 23.82% | accuracy | Verified | Mar 5, 2026 | Details |
| FrontierMath Tier 4 | Reasoning | 37.5% | accuracy | Official | Mar 5, 2026 | Details |
| Vectara Hallucination Leaderboard | Other | 8.3% | Hallucination Rate | Official | Mar 5, 2026 | Details |
| MultiNRC | Reasoning | 62.27% | accuracy | Official | Mar 5, 2026 | Details |