Track what Brave Search's AI Mode says about your brand and which sources it cites, returned as structured JSON you can diff over time.
I kept asking myself a question a spreadsheet could not answer: what does Brave Search's AI Mode actually tell people who look up my product? Brave runs an independent index with a privacy-minded audience, so its AI answer is not a copy of Google's. To watch it properly I needed the answer as data, not a screenshot, so I built the Brave AI Mode API on Apify.
The Brave AI Mode API is a Brave AI answer scraper: give it a query and it returns the AI-generated answer, the sources it cites, and the supporting web results as one structured JSON record.
The AI answer text for your query
The references it cites, with title and URL
The supporting web results beneath the answer
The country and language of the search
Open the Brave AI Mode API and click Try for free.
Enter a query, or a queries list for a batch, plus optional country and language.
Run it and download the answers as JSON or CSV.
Every run is reachable over the Apify REST API, so you can fold Brave answer monitoring into a scheduled job.
POST /v2/acts/johnvc~brave-ai-mode-api/runs
Content-Type: application/json
{ "query": "best password manager", "country": "us", "language": "en" }Registered through Apify's MCP server, the Actor becomes a tool that Claude, Claude Code, and Cursor can call. Ask your agent what Brave's AI says about a topic and which domains it cites, and it runs the scraper and hands back structured rows, no code.
The examples page has runnable recipes: a Brave Search API alternative that returns AI answers as JSON, batch checks across many keywords, brand-mention tracking, and cited-source tracking. The full listing lives on the Peerlist project.
Each record captures Brave's live AI Mode answer surface for your query: the generated answer, the citations under it, and the supporting web results. Answers are produced fresh, so track the claims and citations rather than exact wording, since phrasing shifts between runs even when the substance holds.
Post your query to the Actor's run endpoint and read the dataset when the run finishes. One query resolves to one JSON row with the answer, citations, and results, and you can export the dataset as CSV for a spreadsheet.
Yes. Over Apify's MCP server it works from Claude, Claude Code, and Cursor with no Brave API key to provision, so your agent can consult Brave's AI Mode mid-conversation.
About a cent and a half per query resolved, covering the answer, its citations, and the supporting results in one record. There is no subscription, and the free credit on a new Apify account funds a first monitoring batch.
Yes. The Brave AI Mode API supports agentic payments via the x402 protocol, so AI agents and MCP clients can pay for runs in USDC on Base with no Apify account or API token. Point your agent at the Apify MCP server and it can discover, pay for, and run the scraper autonomously; the Apify x402 announcement has the details.
I answer issues on the Actor's page within a day, so if a query type stops returning an answer, tell me and I will fix the parser.
1
2
0