The tweet was deleted by the author.
But we saved everything 🙂.
OpenAI has announced the release of GPT-5.6, which delivers a significant improvement in design judgment capabilities.
The latest version enhances the AI model's ability to inspect and refine its rendered results by utilizing stronger computer-use skills. This means GPT-5.6 can now evaluate visual and functional aspects of its output, not only generating code or content but catching errors and applying final adjustments before delivering results to users. Early feedback points to improved reliability and higher quality in the completed work.
View post on X Loading...
Earlier this week, OpenAI broadened worldwide preview access to GPT-5.6 Sol by launching it with Terra and Luna, according to a company statement (OpenAI launches GPT-5.6 Sol). In recent months, the company also integrated agentic tools and Codex across its operations to support more complex projects (OpenAI expands internal Codex usage). These steps precede the standard rollout of new design judgment features in GPT-5.6.