Tapesearch

Podcasts contain valuable spoken information, but that content is locked inside hours of audio with no easy way to search it. Market researchers, financial analysts, journalists, advertisers and academics who want to track brand mentions, sentiment or expert commentary currently have to listen through entire episodes to find relevant moments, which wastes hours of 'audio trawling'. People who prefer reading to listening, or who are Deaf or hard of hearing, are effectively excluded from this content entirely. Podcasters themselves struggle to repurpose their own back catalog without searchable transcripts.

Tapesearch indexes a claimed 5 million transcripts and lets users search text within them, with filters, and jump straight to the exact timestamped audio clip. It offers an AI Chat feature to ask questions across multiple episodes and get synthesized answers, email alerts (up to 50 on higher tiers) when tracked keywords are mentioned in new episodes, and trend charts showing how topics evolve across the podcast landscape over time. Transcripts can be downloaded in TXT, SRT, VTT and JSON formats, and an API plus hosted MCP server (compatible with Claude and other AI assistants) lets developers build the data into their own apps, with up to 2000 requests per day. The database refreshes daily, with podcasts on the Podping notification system transcribed within minutes of publishing.