Free Tool · BYOK
TL;Listen
Turn articles and videos into audio briefings
Paste any article URL or YouTube link. TL;Listen reads it, distills the key points, and generates a short podcast-style audio briefing you can listen to while commuting, walking, or doing dishes.
Launch App →Paste any link
Articles, blog posts, YouTube videos, newsletters. If it has text, TL;Listen can summarize it.
Audio briefing
Get a concise, podcast-style summary generated in seconds. Listen on the go instead of reading at your desk.
Bring your own keys
Uses your Claude and ElevenLabs API keys. Optionally set a custom Voice ID. No signup needed — keys stay in your browser, never on our servers.
Who TL;Listen is for
- Commuters and walkers who save 20+ articles a week and never find time to read at a desk.
- Researchers and analysts who want to skim long reports and video explainers without watching at 1.75× with your eyes glued to a window.
- Creators and operators who read to stay ahead and would rather get through 3 articles on a walk than 1 at lunch.
- Accessibility-first readers who prefer audio to long-form text for fatigue, dyslexia, or vision reasons.
Not for you if: you want a full AI podcast host or persistent library. TL;Listen generates briefings on demand — it does not store your queue on a server, and there's no feed or subscription layer.
How TL;Listen is different
The “article to audio” category is crowded. Here's what sets this one apart.
Paste-any-link input, not just articles
Feed it a blog post URL, a Substack essay, a Medium piece, or a YouTube video. TL;Listen extracts the text (or transcript), summarizes with Claude, and voices it with ElevenLabs. One tool, multiple source formats.
BYOK pricing, not per-minute
No $10/month subscription that caps your listening time. Bring your own Anthropic and ElevenLabs keys, pay per-use at their rates. For most readers this is pennies per briefing.
Choose your voice
ElevenLabs voice library at your fingertips, and an optional custom Voice ID field if you've cloned your own. Your briefings sound like the voice you like, not a generic TTS robot.
Nothing stored server-side
No account, no library, no content logs. Your API keys live in browser storage, your source text flies straight from your browser to Anthropic and ElevenLabs. TwoSetAI has no backend for this app.
Frequently asked questions
What is TL;Listen?
TL;Listen is a free browser-based tool that turns articles, blog posts, and YouTube videos into short podcast-style audio briefings. Paste a link, pick a voice, and get an audio summary you can play in the browser or download for the commute. It uses Anthropic Claude to summarize and ElevenLabs to generate the voice.
What sources can TL;Listen read?
Article URLs (blog posts, Substack, Medium, news sites), YouTube video links (transcripts are fetched automatically), and any page with extractable body text. If the content is behind a paywall or requires JavaScript to render, extraction may fall back to the page's metadata.
How much does TL;Listen cost?
The app is free. You bring your own API keys for Anthropic Claude (summarization) and ElevenLabs (voice), and pay those providers directly at their published per-token and per-character rates. A typical 3-minute briefing costs a few cents in combined API spend.
What voices can I use?
TL;Listen uses your ElevenLabs account, so any voice available to your ElevenLabs tier is available to you — including the full public voice library. If you've cloned your own voice in ElevenLabs, paste your Voice ID into the optional custom-voice field to narrate briefings in your own voice.
Is my data stored on TwoSetAI servers?
No. API keys, URLs, and generated audio live in your browser. Source content is sent directly from your browser to the Anthropic and ElevenLabs APIs using your own keys. TwoSetAI does not operate a backend for TL;Listen, so there's no account, no queue, and no logs on our side.
How long is a typical briefing?
Claude targets a short commute-length audio briefing for most article-length inputs. Longer sources get a proportionally longer summary, but TL;Listen optimizes for commute-length audio rather than a word-by-word read-through.
How is TL;Listen different from other read-aloud apps?
Most read-aloud apps read text verbatim or word-for-word. TL;Listen summarizes first, then voices the summary. That trade-off — fidelity vs. brevity — makes it better when you want the point in a few minutes, worse when you want the full text narrated. Plus TL;Listen is BYOK and has no account or queue.
Does TL;Listen work on YouTube videos without transcripts?
TL;Listen fetches the auto-generated transcript YouTube provides for most videos. Videos with transcripts disabled or with heavily non-verbal content (music, silent demos) won't work well. For those, try dropping in a related article link instead.
Who built TL;Listen?
Angelina Yang, the sole operator of TwoSetAI and a fast.ai fellow. Built it after realizing she was saving articles she'd never come back to read. Part of Angelina's Lab.