Particle's New Radar Lets You Search Every Podcast

Particle, the AI news company founded by former Twitter engineers, just made a big pivot. On Wednesday it introduced Radar, a search engine for podcasts. Radar doesn't just transcribe audio—it understands what people are talking about. It indexes more than 130,000 podcasts, including every Apple Top 200 show across 135 categories, and adds 20,000 new episodes every day. That makes it the largest transcribed podcast library in existence.

Why does this matter to you? Because podcasts have been a black hole for search. You can't Google what someone said on a show unless it's manually transcribed. Radar changes that. You can search for any topic, person, company, or product, and it pulls up the exact moment someone talks about it. It even extracts short, self-contained clips with timestamps, so you can listen to just the part you care about. You can also set alerts—get an email or Slack message whenever a guest mentions a specific topic on any podcast.

The real money, though, is in Radar's API, which lets AI agents (software that can take actions for you) tap into this audio intelligence. Until now, AI systems mostly read text from websites; they were blind to audio. Radar gives them ears. That's why hedge funds are already paying for it—they want to find non-public investment signals buried in podcasts. Marketing teams can also use a special ads search engine to see where brands advertise and track how often they appear.

Radar isn't cheap for individuals: $29 per month per seat, with business plans starting at $399 a month. But if you're a researcher, journalist, or just a podcast fan who hates wasting 40 minutes on a dull episode, it could be a game-changer. Next up, Particle plans to expand beyond podcasts to YouTube videos and news clips.