Select a platform above
Welcome back
0
words
Essay type
Review of
Document type
Platform
Occasion
Job match
File metadata
Link detected
What are you checking?
…or paste anything — even a link to a post or article.
Drop your file here
Rewrite
Original
Rewritten
Social Post
Scanning vocabulary...
Matching patterns...
Analyzing structure...
Scoring signals...
Done!
--
--
Essay type
Review of
Document type
Occasion
Platform
Your draft
Flagged — your coach can see these too
Your editor
Image Authenticity Check
Upload or drop an image to check if it's real, AI-generated, or manipulated.
Drop an image here or click to upload
PNG, JPG, WEBP up to 5MB
Uploaded image
Video Slop Check
Analyze any YouTube video for clickbait, AI slop, and transcript quality.
YouTube: click ... → Show transcript → copy
Adjusted for spoken content
Audio Slop Check
Paste a podcast link, upload audio, or paste a transcript — get a TL;DR, fact-checks, and a slop score.
Drop an audio file here or click to upload
MP3, WAV, M4A, OGG
Adjusted for spoken content
How Audio analysis works
Three ways in
• Paste a podcast / MP3 link → Transcribe & TL;DR
• Drop or upload an audio file (MP3, WAV, M4A, OGG)
• Paste a transcript directly → Analyze Transcript
What you get
• A TL;DR summary + key points
• Fact-checks with sources; low-value personal claims tuck under “Show more”
• A spoken-content slop score — casual grammar and filler are not penalized
• Speaker labels (Speaker 1 / 2…) when the audio has more than one voice
The buttons
• Clean up — fixes speech-to-text mishearings (a name or foreign word) from context; never rewrites your words.
• Split speakers — if a transcript came out as one block, AI splits it into speaker turns.
0:00
How to use Live Analysis
What Live is for
Two ways in: paste a link for an instant TL;DR + fact-check without watching, or capture a running voice live — streams, meetings and calls, debates, lectures, podcasts. If it's playing in a tab, Live can listen to it.
The controls
• TL;DR — instant summary + fact-check + slop score of the pasted link (no watching), or of the transcript you've already captured.
• Embed — embed the video so you can capture and analyze it live.
• Speed (on the embedded video) — play a long video faster to capture it in less time. Pitch is preserved, so transcription stays accurate up to 2× — a 1 hr talk becomes ~30 min of capture at half the cost.
• Source (Microphone / Browser tab audio) — where the audio comes from.
• Slop + facts — what to analyze: slop + fact-checks, fact-checks only, or plain transcript.
• + Speaker — mark that a new person started talking.
• Split speakers — let AI split the merged transcript into separate speaker turns.
• Report — download the transcript, scores, and fact-checks.
• Start / Stop — begin or end a live session.
Capturing audio (no extension needed)
A video or tab (best): source = Browser tab audio → Start → in the picker choose the tab and turn on "Also share tab audio" → Share. Clean audio, works with your Mac muted.
In person: source = Microphone, turn speakers up if a video is playing, then Start.
Zoom / Teams / Meet: join in the browser and share that tab's audio — cleanest. Desktop app or phone: speakerphone on + Microphone.
Speakers — when to press what
• Auto (usually nothing to press): on Browser tab audio, voices are labeled automatically (Speaker 1, Speaker 2…).
• + Speaker: on the mic, where voices aren't separated — tap it each time a new person starts talking.
• Split speakers: if it came out as one block, run this once and AI splits it into speaker turns after the fact.
What the readings mean
• Score — spoken-content slop, live-estimated: vague, templated, substance-free talk scores high. Casual grammar and filler are normal in speech and NOT penalized.
• Flagged — rhetoric patterns as they happen: hyperbole, deflection, engagement bait, empty superlatives.
• Fact check — AI grades claims against known sources, tagged with the speaker; click a claim to jump to that moment. It hides low-value personal claims by default — use Show all / Key claims only in the header to toggle.
Tips
• Slop updates every 3s, facts every 30s
• A green glow on a button means its result is ready
• Stop pauses the video and runs a final fact-check + TL;DR
Speed
Turn up your speakers — the microphone captures the video audio
Embed a video or click Start to begin transcription
Live score
What this score means
The big number is a live-estimated slop score for spoken language — how templated, vague, or AI-patterned it sounds. It's a streaming meter, not a verdict.
Two separate numbers
• Originality — inverse of the slop score. Higher is more natural and specific.
• Fact score — average accuracy of claims checked so far, unrelated to writing style. Higher is more accurate; shows — until at least one claim has been checked.
How fact-checks affect the slop score
False or misleading claims nudge the slop score up a little. Verified claims don't lower it — accuracy is tracked separately in Fact score, not blended into the writing-style number.
Scores appear here during live analysis
I can summarize, fact-check, or suggest questions about what's being said.