GPT-5.4
di OpenAI · Rilascio 2026-03-05
59.0
punteggio medio
$2.50/1M
Prezzo Input
$15.00/1M
Prezzo Output
1.1M tokens (~525 books)
Finestra di Contesto
multimodal
Tipo
Tested on 16 benchmarks with 59.0% average. Top scores: Chatbot Arena Elo — Overall (1465.8%), OTIS Mock AIME 2024-2025 (95.3%), ARC-AGI (93.7%).
Punteggi Benchmark
| Benchmark | Categoria | Punteggio | Bar |
|---|---|---|---|
| Chatbot Arena Elo — Overall | arena | 1465.8 | |
| OTIS Mock AIME 2024-2025 | math | 95.3 | |
| ARC-AGI | reasoning | 93.7 | |
| GPQA diamond | knowledge | 91.1 | |
| SWE-Bench verified | coding | 76.9 | |
| ARC-AGI-2 | reasoning | 74.0 | |
| Artificial Analysis — Agentic Index | speed | 69.4 | |
| WeirdML | coding | 57.4 | |
| Artificial Analysis — Coding Index | speed | 57.3 | |
| Artificial Analysis — Quality Index | speed | 57.2 | |
| FrontierMath-2025-02-28-Private | math | 47.6 | |
| SimpleQA Verified | knowledge | 44.8 | |
| Chess Puzzles | knowledge | 44.0 | |
| APEX-Agents | agentic | 35.9 | |
| FrontierMath-Tier-4-2025-07-01-Private | math | 27.1 | |
| PostTrainBench | knowledge | 20.2 |
Modelli Simili
U
Muse SparkUnknown
59.0
DeepSeek
59.0
Google DeepMind
59.1
Microsoft
58.6