Exa Adds 3.2 Million Podcasts to Search Index

Exa announced on October 29, 2026, that its search index now includes 3.2 million podcast episodes from 119,000 shows, with full transcripts. The company says the collection contains roughly 23.7 billion words, or about 2.2 million hours of audio, and is available to AI agents across all Exa endpoints at no additional charge. Exa says it updates the index constantly, with new episodes appearing soon after they air.

The index covers technology, business, finance, news, science, culture, sports and other topics. Search results can include show and episode titles, publisher, description, hosts and guests, publish date, duration, and full transcripts with speaker labels. Exa says this lets users search what was said in episodes, including what Jensen Huang said about inference demand, or find discussions of pricing mistakes.

Exa benchmarked podcast search across 7 APIs, using each API's best tested mode. In a test of 250 active shows, Exa returned the newest episode in the top 5 results 98.4% of the time, higher than the alternatives tested. In a separate test using 667 questions about specific episodes, it produced correct answers 85.0% of the time, also higher than the alternatives tested.

Exa lists personal assistants, investment and market research, sales and marketing research, and recruiting as possible uses. Analysts can track podcast comments about demand, pricing or competition between earnings calls, while agents can retrieve quotes with the speaker and date. Marketers can use transcripts to score accounts and leads. Data partner Particle contributes to the index, which is live in Exa's search API for all users at no additional cost.

Exa announced on October 29, 2026, that its search index now includes 3.2 million podcast episodes from 119,000 shows, with full transcripts. The company says the collection contains roughly 23.7 billion words, or about 2.2 million hours of audio, and is available to AI agents across all Exa endpoints at no additional charge. Exa says it updates the index constantly, with new episodes appearing soon after they air.

The index covers technology, business, finance, news, science, culture, sports and other topics. Search results can include show and episode titles, publisher, description, hosts and guests, publish date, duration, and full transcripts with speaker labels. Exa says this lets users search what was said in episodes, including what Jensen Huang said about inference demand, or find discussions of pricing mistakes.

Exa benchmarked podcast search across 7 APIs, using each API's best tested mode. In a test of 250 active shows, Exa returned the newest episode in the top 5 results 98.4% of the time, higher than the alternatives tested. In a separate test using 667 questions about specific episodes, it produced correct answers 85.0% of the time, also higher than the alternatives tested.

Exa lists personal assistants, investment and market research, sales and marketing research, and recruiting as possible uses. Analysts can track podcast comments about demand, pricing or competition between earnings calls, while agents can retrieve quotes with the speaker and date. Marketers can use transcripts to score accounts and leads. Data partner Particle contributes to the index, which is live in Exa's search API for all users at no additional cost.