Body (Markdown)

## A New Era of Multimodal AI

OpenAI today unveiled GPT-5, the latest iteration of its flagship large language model, marking a significant leap in multimodal reasoning capabilities.

The model demonstrates unprecedented ability to chain complex reasoning across text, images, and code simultaneously — a capability the company calls "unified cognition."

### Key Improvements

- **Cross-modal reasoning**: GPT-5 can analyze an image, write code to process it, and explain the results in natural language — all within a single inference pass
- **Extended context**: 1M token context window as standard, with 4M available on API tier
- **Latency**: 40% faster than GPT-4.5 on equivalent tasks

> "This isn't just a bigger model. It's a fundamentally different architecture," said Sam Altman during the announcement.

### Pricing and Availability

The model is available immediately via API at $15/M input tokens and $60/M output tokens. ChatGPT Plus subscribers will get access within the next two weeks.

Edit Article