Particle, the AI newsreader startup founded by former Twitter engineers, launched Radar on Wednesday, a podcast search engine and API that transcribes more than 130,000 shows and lets AI agents query the spoken word the way they already query the web. The service covers all 200 podcasts in the Apple Top 200 across 135 verticals and ingests 20,000 fresh episodes every day. Radar is priced at $29 per month per seat, with a $399 per month business tier that includes 20 seats, and custom pricing for direct API integrations.
The pitch is straightforward: audio has been a blind spot for the entire agent economy. Web crawlers, retrieval systems, and search APIs are built for text, which means the millions of hours of executive interviews, analyst commentary, and product discussion happening inside podcasts every week are effectively invisible to any autonomous system trying to reason over the news. Radar transcribes that audio, tags speakers, and enriches each transcript with entity metadata covering people, companies, brands, products, and topics.
Particle CEO Sara Beykpour told TechCrunch that the company's ambition is to become the default audio layer for agent infrastructure, not just a consumer podcast tool. The company had originally built audio-clip surfacing as a feature inside its news-reading app, then realized the underlying capability was more valuable as a standalone API as the agent market took off.
Key facts
- 01Radar transcribes more than 130,000 podcasts, including every show in the Apple Top 200 across 135 verticals.
- 02The service adds 20,000 new episodes to its index every day, with speaker labels and entity metadata.
- 03Pricing starts at $29 per month per seat, with a $399 per month business plan that includes 20 seats.
- 04Hedge funds are the highest-volume API customers, integrating podcast intelligence directly into their agent stacks.
- 05AI search API provider Exa is among Radar's launch partners.
The early revenue signal is coming from finance. Hedge funds are integrating Radar's API directly to feed their models with data they can't get from text-only sources — earnings-adjacent commentary from CEOs on interview podcasts, sector chatter from industry shows, mentions of specific tickers or products across thousands of episodes a week. AI search platforms and data resellers make up the other high-value segment. Exa, the AI-agent-focused search API, is among the launch partners.
On top of raw transcripts, Radar tracks mentions of entities across the full podcast corpus and pushes alerts via email, Slack, or webhook whenever a target person, company, or topic surfaces. Users can filter by guest, topic, or podcast tier — say, only alert when a named executive appears on a top-ranked show — and receive results in real time or as daily and weekly digests.
“Hedge funds have been the highest-volume customers that are directly integrating with the API.”— Sara Beykpour, Particle co-founder and CEO
The product also extracts self-contained clips with timestamps, so users can jump straight to the moment a topic is discussed rather than skim a summary. Beykpour said Particle pre-selects notable clips to give listeners a fast way into an episode without committing to the full hour. Additional layers track ad reads, listener ratings, chart rankings, audience-size estimates, and brand-suitability signals — each of which opens a separate monetization path against advertisers, agencies, and sponsorship analytics buyers.
There is a dedicated podcast ad search engine inside Radar that returns every episode where a given company has advertised and charts how that spend trends over time. For brands trying to audit their own podcast media mix or scout competitor placements, that dataset has no direct equivalent in text-based ad intelligence tools. Political bias analysis and sponsorship data round out the intelligence layer.
Radar's roadmap extends beyond podcasts. Particle plans to add YouTube videos and news clips to the index, positioning the API as a general audio-and-video intelligence service rather than a podcast-only tool. The direction lines up with where the agent stack is heading — as more autonomous systems move from answering questions to monitoring the world, the ones that can read spoken sources will operate on a materially larger information surface than the ones that can't.
The business risk is that the biggest audio platforms — Apple, Spotify, YouTube — could ship comparable transcription and search features natively and undercut a third-party API on price. Particle's bet is that agent developers and hedge funds want a neutral, cross-platform index with programmatic access and entity-level metadata, not a walled-garden search box tied to one distributor. Whether that neutrality is worth $29 a seat or six-figure API contracts will depend on how quickly the incumbents move.
Radar is the kind of infrastructure play that only makes sense once agents become buyers. A podcast search tool sold to humans is a niche product; a podcast intelligence API sold to autonomous systems that need to monitor thousands of hours of audio a day is a category. Particle is betting the agent economy is now real enough to support a dedicated audio layer underneath it — and the fact that hedge funds are already the highest-volume customers suggests at least one segment has already answered yes.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
