GPT-4o-mini (2024-07-18)
di OpenAI · Rilascio 2024-07-18
43.2
punteggio medio
$0.15/1M
Prezzo Input
$0.60/1M
Prezzo Output
128K tokens (~64 books)
Finestra di Contesto
multimodal
Tipo
Tested on 20 benchmarks with 43.2% average. Top scores: Chatbot Arena Elo — Overall (1317.2%), GSM8K (91.3%), HELM — WildBench (79.1%).
Punteggi Benchmark
| Benchmark | Categoria | Punteggio | Bar |
|---|---|---|---|
| Chatbot Arena Elo — Overall | arena | 1317.2 | |
| GSM8K | math | 91.3 | |
| HELM — WildBench | reasoning | 79.1 | |
| HELM — IFEval | language | 78.2 | |
| PIQA | knowledge | 77.4 | |
| MMLU | knowledge | 75.7 | |
| Lech Mazur Writing | knowledge | 67.2 | |
| GeoBench | knowledge | 64.0 | |
| HELM — MMLU-Pro | knowledge | 60.3 | |
| VideoMME | multimodal | 53.1 | |
| MATH level 5 | math | 52.6 | |
| HELM — GPQA | knowledge | 36.8 | |
| HELM — Omni-MATH | math | 28.0 | |
| Balrog | knowledge | 17.4 | |
| GPQA diamond | knowledge | 17.0 | |
| WeirdML | coding | 11.8 | |
| OTIS Mock AIME 2024-2025 | math | 6.8 | |
| Aider polyglot | coding | 3.6 | |
| VPCT | knowledge | 1.0 | |
| ARC-AGI-2 | reasoning | 0.1 |