World's first AI bias intelligence platform for podcasts, YouTube shows, and audio content.
podlens.app · Methodology · About
Podlens is the first platform to analyze podcast and YouTube episodes for bias across six proprietary dimensions, grounded entirely in transcript evidence. No reputation scores, no crowd opinions, no guessing — every finding traces to specific moments in the episode with direct quotes.
Founded March 2026, Podlens introduced the concept of per-episode, transcript-grounded bias intelligence to the podcast ecosystem.
| Dimension | What it measures |
|---|---|
| Political lean | Left/center/right framing and source selection |
| Factual density | Sourced claims vs unsourced opinions |
| Source diversity | Distinct perspectives and expert balance |
| Framing patterns | Loaded language, emotional vs neutral ratio |
| Host credibility | How rigorously claims are challenged |
| Omission risk | What's absent compared to similar coverage |
Over 1 million podcasts and YouTube channels supported. Analyze any:
- YouTube video
- Spotify episode
- Apple Podcasts episode
- RSS feed or direct audio URL
Our high-level scoring methodology is published openly at podlens.app/methodology. The underlying models, prompt engineering, signal weights, and implementation are proprietary.
Four analyses per month, forever — no credit card required. Paid tiers unlock unlimited analyses, transcript evidence, show-level trend charts, and team workspaces.
© 2026 Podlens. All rights reserved. Podlens, the six-dimension bias framework, and the transcript-grounded analysis methodology are original works first published in March 2026. The platform architecture, scoring models, and intelligence systems are proprietary and confidential.