GPT-5.4 Pro places in the top three on 8 benchmarks.
Closed weights
Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| MultiNRC | 62.3% | #4 of 38 | Official leaderboard | |
| Humanity's Last Exam | 44.3% | #6 of 47 | Official leaderboard | |
| SkateBench v2 | 78.5% | #3 of 21 | Official leaderboard | |
| MultiChallenge | 69.2% | #9 of 35 | Official leaderboard | |
| Vectara Hallucination Leaderboard (2025 dataset) | 8.30% | #42 of 104 | Official leaderboard | |
| SimpleQA Verified | 46.3% | #38 of 84 | Epoch AI |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| GPQA Diamond | 94.6% | #7 of 219 | Epoch AI | |
| FrontierMath | 50% | #3 of 70 | Epoch AI | |
| FrontierMath Tier 4 | 37.5% | #3 of 55 | Epoch AI | |
| CritPt | 30% | #10 of 160 | Epoch AI | |
| MathArena (Apex, IMO, USAMO) | 69.8% | #3 of 47 | Official leaderboard | |
| SimpleBench | 74.1% | #9 of 82 | Epoch AI | |
| FrontierMath Tiers 1-3 v2 | 82.5% | #15 of 89 | Epoch AI | |
| ARC-AGI-2 | 83.3% | #17 of 95 | Official leaderboard | |
| ARC-AGI | 94.5% | #19 of 92 | Official leaderboard | |
| FrontierMath Tier 4 v2 | 58.5% | #18 of 71 | Epoch AI | |
| FrontierScience-Research | 36.7% | #1 of 3 | Model card | |
| xbench-ScienceQA | 72.2% | #11 of 29 | Official leaderboard |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| WeirdML | 57.4% | #35 of 130 | Epoch AI |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Chess Puzzles | 59% | #9 of 83 | Epoch AI | |
| BrowseComp | 89.3% | #10 of 77 | Model card | |
| PinchBench | 12% | #47 of 48 | Official leaderboard |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| prinzbench | 79.0 | #3 of 26 | Official leaderboard | |
| TaxCalcBench | 62.8% | #2 of 17 | Official leaderboard | |
| Vals Finance Agent | 61.5% | #3 of 8 | Model card | |
| GDPval | 82% | #5 of 10 | Model card | |
| Investment Banking Modeling Tasks (OpenAI) | 83.6% | #2 of 4 | Model card |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| VISTA | 53.9% | #3 of 54 | Official leaderboard | |
| EnigmaEval | 23.8% | #5 of 41 | Epoch AI |
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| FORTRESS | 14.8% | #13 of 66 | Official leaderboard | |
| Human Pathogen Capabilities Test v2 (HPCT v2) | 63.9% | #20 of 71 | Official leaderboard |
OpenAI's models in color, against every other model on the Epoch Capabilities Index, by release date.
The benchmarks the lab chose to headline, with the scores it gave.