Visual mode vs Deep mode

Divrite offers two levels of research. Which one you pick changes how many sources are read, how long the run takes and how many credits it uses.

For questions: Visual or Deep

Visual runs a broad web search and writes an answer that includes original images from the pages it read, plus charts built from the data it found. Each image keeps a link to the page it came from. It usually finishes in about a minute.

Use Visual when you want a quick, well-sourced overview: what something is, what happened recently, how a few options compare.

Deep treats your question as a research project. It splits the question into sub-topics, searches them in parallel, checks which points are still unclear, and runs follow-up searches on those points before it writes. The result is a longer, sectioned report. It usually takes a few minutes.

Use Deep when the answer has to hold up: policy and regulation, market and competitor research, due diligence, or anything with several moving parts.

For presentations and documents: research depth

When you create a presentation or document, you choose a research depth:

• Deep research splits the topic into tasks, then searches and cross-checks each one. Expect roughly 10–15 minutes.
• Quick web search runs one round of web search and then writes. Expect roughly 3–5 minutes.
• Decide later lets you start now and choose once the task begins.

Times vary with the topic and how much material there is to read.

Which should I pick?

• If you’ll present or publish the result, or the numbers matter, pick Deep.
• If you’re exploring a topic or need a fast first draft, pick Visual or Quick web search, then go deeper on the parts that matter.

Deeper research reads more sources, so it uses more credits. See How credits work for details.