Deep Search: real-time web research built into your chat

Research used to mean opening fifteen tabs, reading each one, copying quotes into a doc, and then — finally — writing.

Deep Search compresses all of that into a single chat message.

How it works

When you enable Deep Search in the chat toolbar, QuBold:

1. Queries live web sources in real time
2. Reads and extracts relevant content from top results
3. Synthesizes a structured answer with inline citations
4. Saves the sources to a notebook attached to your conversation

You get a grounded, cited response — not a hallucinated one.

When to use it

**Competitive research:** "What are the main criticisms of [competitor] on Reddit and review sites right now?"

**Current events:** "What happened with [company] this week?"

**Fact-checking:** "Is this claim about [topic] accurate based on recent reporting?"

**Market sizing:** "What do analysts say about the TAM for AI chatbots in 2026?"

Deep Search is particularly useful for anything time-sensitive — training data cutoffs don't affect it.

What gets saved

Every Deep Search response auto-saves: - The query and AI summary - Cited URLs with extracted quotes - A structured notebook entry you can reference later

Your research stays organized without any extra effort.

Deep Search vs. standard chat

Standard chat uses the model's training knowledge — fast, good for stable facts and reasoning tasks. Deep Search adds live retrieval — slightly slower, essential for current information.

Use standard chat for drafting and brainstorming. Switch to Deep Search when you need to know what's true *right now*.