OpenAI Launches GPT-5: Reasoning Capabilities Surpass GPT-4o Across the Board, Enterprise Edition Also Available

OpenAI officially released the GPT-5 model, which comprehensively surpasses GPT-4o on benchmarks for math, science, and programming, with reasoning capabilities improved by approximately 40%. The enterprise edition is also available, supporting a 128K context window and enhanced tool-calling capabilities. API pricing is about 30% lower than GPT-4o, significantly reducing development costs for enterprise AI applications.