OpenAI Launches GPT-5 with Enhanced Reasoning and Multimodal Capabilities

TL;DR: OpenAI has officially released GPT-5, featuring significant improvements in reasoning, a 1M token context window, and native vision-language capabilities.

Key Points

• GPT-5 achieves state-of-the-art results on major reasoning benchmarks, outperforming GPT-4o by 40% on GPQA.
• The model supports a 1 million token context window natively, up from 128K in GPT-4o.
• Native multimodal support allows processing of images, audio, and video in a single inference call.
• API pricing starts at $5 per million input tokens, roughly 2x the cost of GPT-4o.
• Available immediately through the API and rolling out to ChatGPT Plus subscribers over the next week.

Why It Matters

GPT-5 represents a meaningful step forward in AI reasoning capabilities. The expanded context window opens up use cases in document analysis, codebase understanding, and long-form content generation that were previously impractical. For AI tool builders, the improved reasoning could enable more reliable autonomous agents and complex multi-step workflows.

Source: OpenAI Blog