Particle's Radar makes podcasts searchable and machine-readable

Particle, the startup founded by former Twitter engineers, is abandoning its news-reader app for what could be a far more lucrative direction: indexing the spoken conversations buried in podcasts and turning them into data that machines can consume. On Wednesday the company launched Radar, a podcast search engine that doesn't just transcribe audio — it understands the meaning, and can surface key quotes and highlights from the conversation.

The numbers the company publishes put it well ahead of competitors: more than 130 thousand podcasts already transcribed, making Radar the largest transcription service of its kind. The index covers all 200 of Apple's top podcasts across 135 categories, with 20 thousand new episodes added daily. Transcripts include speaker diarization and rich metadata, thanks to entity recognition that picks out people, companies, brands, products and topics as they come up in conversation.

The system can track entity mentions across podcasts and send real-time alerts, or a daily or weekly digest, via email, Slack or webhook. Filters let users narrow alerts to specific guests speaking on a defined topic, or restrict search to top-tier podcasts only. Radar also extracts standalone clips with timestamps, so users can both listen and read the relevant segment. "We pre-select notable clips, so if you don't have time to listen to a full episode and don't want to read a summary, this is the fastest way to understand what's happening there," explained CEO and co-founder Sara Beykpour.

Beyond the web interface, the real product is the API and MCP, which let AI agents and businesses access that same intelligence programmatically. Hedge funds are the highest-volume customers embedding the API directly, Beykpour said — they're looking for data their agents can't see on the open web. Other paying customers include AI search platforms and data brokers; agent-search API provider Exa is named as a partner. The vision, she said, is to centralize all new-media and audio intelligence in a single API, because most agents today are blind to audio unless someone has transcribed it for them.

Pricing starts at $29 per month per seat, with a business plan at $399 per month for 20 seats; API customers receive custom pricing. The system also offers a dedicated search engine for podcast ads, locating every episode where a given company advertises and tracking trends over time, alongside political-bias analysis, chart rankings, audience-size estimates, sponsorship data and brand-fit scoring — all additional monetization layers on top of basic search.