The new state-of-the-art image model produces precise, immediately usable visuals.

OpenAI calls it a “step change” in instruction following, text rendering, and multi-aspect-ratio generation.

The model can also search the web for real-time context with thinking mode (paid plans only).

It’s now available in ChatGPT, Codex, or the API.

See what it can do here.

_{Source: Superhuman AI}

Here are some resources that other subscribers love: