JEVMETER
CLI that transcribes a video, has Jev score every sentence against rubric flags, and renders a 16:9 edit with a live meter.
Overview
JEVMETER puts a live Jev meter on any video. It transcribes the speech, asks Jev yes/no rubric questions about every sentence, and renders an edited 16:9 video with a meter overlay, flags on notable sentences, a camera that follows the speaker and a final scoreboard. It started with a 'BS meter' on a US presidential debate, shared on X.
A terminal wizard guides non-programmers: drop in a video, pick a preset, optionally provide a transcript with speakers, and choose a highlights cut or a full-length overlay. Presets cover debates and interviews (evasive, dodged the question, emotional appeal, contradicts self), earnings calls (vague guidance, blaming outside factors, hype, dodges), podcasts (unsupported claims, overgeneralisation, outrage, self-promotion) and pitches (buzzwords, overpromises, urgency pressure, vague benefits). A doctor command checks Python, ffmpeg, speech-to-text and the key. The repository links an eval results file.
What the decision model decides
| Type | Question | Answers |
|---|---|---|
| Yes / no | Is this sentence evasive / does it dodge the question? | — |
| Yes / no | Does this sentence make an unsupported claim or overpromise? | — |
How it works
- Provide a video file or link and choose a rubric preset
- Speech is transcribed and split into sentences
- Jev scores each sentence against the preset's yes/no flags
- Scores drive a live meter and flags in a rendered 16:9 edit
- Output saved next to the original as name.jevmeter.mp4
Strengths
- No-code wizard for non-developers
- Presets for several genres of talk
- Key stored in user settings, not the repo folder
- Doctor command for setup checks
Caveats
- Scores rhetoric labels, not factual accuracy
- Requires ffmpeg, speech-to-text and a TypeSafe key
- Windows needs a manual install
Getting started
git clone https://github.com/ChetasLua/jevmeter.git && cd jevmeter && ./install.sh, then source .venv/bin/activate and run jevmeter.At a glance
- Type
- App
- Status
- beta
- Language
- Python
- Runtime
- CLI with ffmpeg
- Has code
- yes
- License
- MIT
Use cases and integrations
How curated lists describe it
Live Jev meter on any video: every sentence scored, rendered as a 16:9 edit. Demo: [Chetaslua](https://x.com/chetaslua/status/2100473581251748216).
AnotiaWang/awesome-decision-modelsCreates edited videos with score meters by asking Jev to rate transcript sentences against selected rubrics.
logicrw/awesome-jev-projectsScores every sentence in a video and renders the result as an overlay.
cobanov/awesome-jevPut a live Jev (TypeSafe) meter on any video: every sentence scored, rendered as a 16:9 edit.
hellogumbo/awesome-jevScores every sentence in a video and renders a live Jev meter as a 16:9 edit.
MrJev/awesome-jev
Frequently asked questions
What does Jev score?
Each transcript sentence against yes/no rubric flags such as evasive, hype or unsupported claim.
Does it fact-check?
No, it flags rhetorical patterns defined by the preset, not truth.
What do I need?
Python 3.9+, ffmpeg, speech-to-text support and a TypeSafe API key.
Related projects
Claude Code plugin that logs your prompts locally and later has Jev score how well you prompt the coding agent over time.
PluginbetaTypeScriptChrome extension that grades readable page sections on ten rubric scores with Jev via Vercel AI Gateway and gives the page an A-E grade.
ProjectbetaTypeScriptCLI and agent skill that checks every line of a document against every rule with Jev and outputs flags for an AI agent to fix.
Agent skillbetaTypeScriptTournament engine that ranks texts with Swiss pairing, Elo and TrueSkill, using Jev or local models to judge each head-to-head duel.
ProjectprototypePythonFree web tool that checks any URL for 35 tells of AI-generated design and returns a slop score in about 250 ms.
DemoproductionAnalytics engine that rates published posts on pre-registered rubrics with Jev and tests ratings against real engagement.
ProjectbetaTypeScript