Announced August 26, 2026, Radar transcribes and indexes 130,000+ podcasts so their spoken content becomes searchable — and finally readable by the AI agents that were blind to audio.
2026-08-27 · by Moe Ameen
Particle, the AI news startup founded by former Twitter engineers and led by co-founder and CEO Sara Beykpour, announced Radar on August 26, 2026. Radar is a podcast search engine and intelligence platform that transcribes podcast audio at scale and turns it into structured, searchable data, exposed through an API and an MCP (Model Context Protocol) endpoint so both people and AI agents can query what was said on air. Beykpour's framing of the gap: "Agents are generally blind to audio; they can't see it unless something or someone has transcribed it."
Particle says Radar has transcribed more than 130,000 podcasts — describing it as the largest transcribed podcast service — and adds roughly 20,000 episodes a day, covering the Apple Top 200 across 135 verticals. On top of the transcripts, Radar layers speaker labels and metadata that identify entities like people, companies, brands, and topics, and it extracts key quotes, highlights, and self-contained clips with timestamps. Users can track when an entity is mentioned and receive customizable alerts by email, Slack, or webhook, and Radar also runs a dedicated podcast-ads search engine with sponsorship data, listener ratings, and brand-suitability signals.
The early demand is coming from finance and search, not creators. Particle has said hedge funds are its highest-volume API customers, mining spoken conversations for signals their agents otherwise can't access, alongside AI search platforms, data resellers, journalists, and researchers. The idea grew out of a feature in Particle's news-reading app, which used podcast clips to enrich related news stories. At launch, Particle listed a Consumer plan at $29/month per seat and a Business plan at $399/month including 20 seats, with API pricing quoted custom; it has signaled plans to expand beyond podcasts into YouTube videos and news clips. Treat the specifics as a launch-window snapshot and confirm details on Particle's own pages.
There are two ways a creator acts on this. The immediate one: it's a live content moment. Drop "Particle just launched Radar — AI agents can finally read podcasts" into [Kompozy](/) Quick Ingest and fan one source into a Text Post, an X thread, a LinkedIn carousel, a captioned Short, and a blog explainer in your own voice, then schedule and publish the set across the eight social platforms plus blog and email while the story is fresh.
The more durable move is about your own show. Radar makes the point that a podcast is now structured, quotable, agent-readable content — so the episodes you already record are worth far more than one audio upload. That's exactly the repurposing job Kompozy is built for: feed a long episode in and Kompozy cuts [Clipped Shorts](/glossary/clipped-short) from the recording with burned-in captions, drafts [carousels via HyperFrames](/glossary/hyperframes) and [quote graphics](/) from the strongest lines, and spins up a [blog article](/) and a newsletter from the transcript — every output governed by one [Persona Brief](/glossary/persona-brief) and pushed out on [Autopilot](/glossary/autopilot). Radar makes audio findable; Kompozy makes sure your own audio shows up everywhere it should, as finished posts rather than a file only Radar's index has read.
Radar is a podcast search engine and intelligence platform announced August 26, 2026. It transcribes and indexes more than 130,000 podcasts and exposes them as structured, searchable data through an API and an MCP endpoint, so people and AI agents can find what was said on air, complete with speaker labels, entity metadata, and timestamped quotes and clips.
It makes spoken audio machine-readable at scale, which means a quotable podcast line can now be indexed and surfaced to AI agents — new distribution beyond live listeners. It also works as a trend-spotting tool, and it underscores that your own episodes are repurposable content, not just an audio archive.
At launch Particle listed a Consumer plan at $29/month per seat and a Business plan at $399/month including 20 seats, with API pricing quoted custom based on usage. Confirm current pricing on Particle's own pages.