#

Key takeaways

• Particle introduced Radar as a podcast search product.

• Radar transcribes audio and extracts quotes and highlights.

• The product offers an API and MCP for AI agents.

• Particle says Radar indexes more than 130,000 podcasts.

• The source notes that coverage, pricing, and language support need independent verification.

What the source reports

TechCrunch reported on August 26, 2026 that Particle introduced Radar. The product is designed to make podcast audio searchable for AI agents. It transcribes audio, extracts quotes, and surfaces highlights.

The source also says Radar offers an API and MCP for AI agents. That suggests the product is built for integration, not only for direct human search. The report frames Radar as a tool for structured access to spoken content.

Particle says Radar indexes more than 130,000 podcasts across 135 verticals. It also adds 20,000 episodes daily. These are the only scale figures provided in the source.

Why searchable audio matters

Podcast audio is useful, but it is harder to search than text. A system that transcribes episodes can make spoken content easier to scan. Quote extraction and highlight generation can also reduce the time needed to review long episodes.

For AI agents, searchable audio can support retrieval tasks. An agent can look for specific phrases, topics, or segments without listening to full episodes. That can make audio content more practical for workflows that depend on fast lookup.

The source does not describe the underlying model, ranking method, or quality controls. It also does not explain how Radar handles errors in transcription. Those details matter, but they are not supplied here.

What is known about coverage

Particle says Radar covers more than 130,000 podcasts. It also says the product spans 135 verticals. The report does not define those verticals, so they should be treated as the company's own grouping.

The source also says Radar adds 20,000 episodes daily. That indicates ongoing indexing activity. It does not, however, show how quickly new episodes become searchable or how completeness is measured.

Because the report is limited, readers should avoid assuming universal coverage. The source does not say that Radar indexes every podcast or every episode. It only provides the company's stated figures.

Integration for AI agents

The mention of an API and MCP is important. It suggests Radar is meant to connect with other systems. That can matter for teams that want audio search inside larger AI workflows.

The source does not explain the API design or the MCP implementation. It also does not say which platforms or tools can use them. Those are operational details that should be checked directly with the product.

For builders, the main signal is simple. Particle is packaging podcast search as a machine-readable service. That makes the product more relevant to automated discovery than to manual browsing alone.

Risks and verification points

The source explicitly says coverage, pricing, language support, and local accessibility should be verified independently. That is the clearest caution in the report. It means readers should not assume the product fits every use case.

The report does not provide pricing. It does not list supported languages. It also does not describe regional availability. Those gaps are material for any team considering adoption.

The source also does not mention editorial controls, rights handling, or content moderation. Since Radar works with podcast audio, those topics may matter in practice. But they are not described in the supplied material, so they remain unknown here.

Morocco relevance

The source reports no Morocco-specific facts. For readers in Morocco, the global lesson is conditional: if you work with spoken content, searchable audio can make review and retrieval faster. Any local fit still depends on independent checks for coverage, language support, pricing, and access.

Bottom line

Radar is presented as a search layer for podcast audio. It combines transcription, quote extraction, highlights, and agent-friendly interfaces. The product's value appears to come from turning spoken content into structured, searchable data.

The report is useful, but it is also limited. It gives scale figures and integration signals, yet leaves key operational questions unanswered. Anyone evaluating the product should verify the missing details before drawing conclusions.