Brave Search adds conversational AI follow-ups

Searching for something online usually follows an annoying loop: you look up a topic, read a quick summary, and then realize you have a follow-up question. On Google, that means wiping the box and starting a brand new search from scratch.

Brave wants to break that friction. On Thursday, the company announced that Brave Search is adding an interactive AI chat mode right beneath its initial search results.

Search once, chat until you get it

Earlier this year, Brave rolled out "Answer with AI," a feature similar to Google's AI Overviews that puts synthesized answers directly above regular search links.

Now, users will see a new "Ask a follow-up question" chat bar sitting directly below that AI summary.

If you search for "Christopher Nolan films," Brave will generate a clean overview of his filmography. Instead of clicking away or retyping a query, you can simply ask, "Which actors appear most in his films?"

The engine understands the context, provides the answer, and surfaces source citations along the way so you can verify where the data comes from.

Under the hood, Brave says the experience runs on a blend of open-source and proprietary in-house large language models. The idea is to merge the retrieval strengths of search engines with the contextual reasoning of chatbots, ending the awkward split between search-first and chat-first tools.

Privacy remains Brave's core pitch here. The company stressed that it does not store queries or profile users. Brave also shared that its search engine now handles over 36 million queries daily, with Answer with AI powering more than 11 million answers each day.

Why it matters

For everyday researchers and casual searchers, this is an immediate quality-of-life win. Most complex questions aren't solved in a single prompt; they require follow-ups, narrowing down options, and checking sources. Doing all of that in one unified tab without opening five different links is genuinely useful.

It also highlights an interesting strategic gap with Google. Google still forces users into entirely separate searches to explore follow-up questions, likely because keeping users clicking traditional result pages preserves ad impressions. Brave doesn't carry that baggage, allowing it to build the actual user experience people want.

We think combining open and internal models is a smart, cost-effective way to challenge incumbent search engines. That said, keeping multi-turn search conversations accurate is notoriously tough, and maintaining high context fidelity without racking up heavy compute bills will be Brave's real long-term test.