Benchmarks
Model performance scores across popular evaluation benchmarks.
GPT-5
90.2
Fuente: -
Claude Opus 4
90.1
Fuente: -
DeepSeek R1
89.1
Fuente: -
Gemini 2.5 Pro
89
Fuente: -
GPT-4o
83.1
Fuente: -
Model performance scores across popular evaluation benchmarks.