Generate images from text prompts
OpenAI's Deep Research, but open
In-browser unified multimodal understanding and generation.