2025-04-16

2건

🇺🇸 o3 (OpenAI) (2025.4.16)

기업 · DemandSphere · 🚀

Reasoning · Closed · 200K ctx · $2/M · $8/M

GPQA 87.7% · SWE 71.7% · MMLU 87% · HE 97% · MATH 97.3% · AIME 88.9%

Flagship reasoning model. 88% ARC-AGI. Three effort levels. Price dropped ~80% from launch.

🇺🇸 o4-mini (OpenAI) (2025.4.16)

기업 · DemandSphere · 🚀

Reasoning · Closed · 200K ctx · $1.1/M · $4.4/M

GPQA 81.4% · SWE 68.1% · MMLU 85% · HE 95% · MATH 97.4% · AIME 92.7%

Best-value reasoning model with vision. Near-o3 math at 2.4x lower cost.