Complexity

The home screen is a single input with the prompt Ask anything. Type a full question rather than keywords, since the engine is built to handle natural language. Answers stream in rather than appearing all at once, which makes longer responses feel quicker than they are.

Before the answer, Complexity shows a row of source cards with a thumbnail, a headline, and a text snippet. The previews come from each page's metadata, so you can judge relevance without opening five tabs. Clicking a card takes you to the original article.

Answers are short paragraphs, not essays, with key entities and phrases highlighted in orange. The highlighting gives you a rough map of what the model pulled from the sources, and it makes scanning faster when you only need one fact.

Every search becomes a session in the left sidebar, labeled with your original question. Clicking one reloads the sources and the answer, so you can pick up a thread days later. It's closer to a reading list than a chat log, and that suits research better than an endless scroll.

A second input at the bottom of an answer takes follow-up questions. Because the session keeps its earlier sources in view, you can narrow a broad question step by step: what happened, who was involved, what changed since. The context stays put instead of resetting with every query.

The home screen lists a handful of trending queries, from tech news to weather and recipes. They work as icebreakers when you want to see how the tool handles a topic before trusting it with your own question. The mix is broad, which is a fair signal of the engine's general-purpose intent.