Particle has launched Radar, a search service that turns podcast conversations into structured data for people and AI agents. The system transcribes more than 130,000 podcasts and adds about 20,000 episodes to its index each day, according to the company.
Radar labels speakers and identifies the people, companies, products and topics mentioned in an episode. Users can search across shows, extract self-contained clips with timestamps, and receive alerts by email, Slack or webhook when selected entities appear. A separate advertising search can track where a company sponsors podcasts over time.
The web interface is only one part of the product. An API and Model Context Protocol connection let businesses and agents query the same index programmatically. Particle says hedge funds are its highest-volume API customers, with AI search providers and data resellers also using the service. The pitch is straightforward: most automated research tools crawl text and cannot use an audio discussion until someone transcribes it.
Individual access costs $29 per month, while a business plan with 20 seats costs $399 per month. API pricing is negotiated separately. Particle plans to add other audio sources, including videos and news clips.