Flows while you listen
The AI takes notes in the background as each speech happens. You watch the round; it builds the flow sheet.
Powered by GPT-4o + Whisper · Tuned to you
AI does the analysis.
You understand the round.
Free AI for Public Forum debate. Flows every speech, breaks the round down clash by clash, and drafts a full ballot — decision, RFD, speaker points, and per-speaker feedback — in about a minute.
Free. No signup. No API key. Just start a round.
Built for two use cases
Parent & lay judges
Watch the round break down clash by clash, review a drafted ballot (decision, RFD, speaker points), and submit the version that actually reflects what happened.
Debaters reviewing mock rounds
See who won each clash and why, plus what each speaker should work on and what the ballot would have looked like — coach-quality review, no coach required.
Pick your event to start a round
PF live · more events comingComing soon
🗳️LD
📜Policy
🎤Speech
🏛️Congress
~60 sec
High-quality flow, analysis, decision
vs. an hour from a human judge or coach
All 8
PF speeches flowed
Constructive through Final Focus
Free
No signup, no API key, no payment
Open the site, start a round
Private
Audio + evidence stay on your device
Nothing uploaded to a DebateJudge server
What you get
Four deliverables. One click.
After the round, hit Generate. Sixty seconds later the AI has flowed every speech, broken the round down clash by clash, and produced a flow sheet, a round summary, per-speaker feedback, and a paradigm-aligned RFD ready to send.
AI flow sheet
FlowEvery speech, every argument, captured automatically. Color-coded by side.
Pro · Constructive
Alex Chen
C1 carbon tax — Smith '24, 30% emissions ↓. C2 jobs — Brookings, +2.4M green jobs. Frames on long-term planet impact.
Con · Rebuttal
Jordan Park
Turn on jobs — Heritage '24, regressive tax kills 1.8M low-income jobs. Drops mechanism on Smith. Frames economy frame.
Round summary
SummaryBoth cases, every clash broken down (who argued what, who dropped what, who won it), and the notable drops — a digest of the whole round in 20 seconds.
Pro case
Carbon tax cuts emissions 30% — long-term planet impact frame.
Con case
Tax kills 1.8M low-income jobs — near-term economy frame.
Clash 1 · Mechanism
Pro extends Smith '24. Con drops mechanism response in CC.
Notable drop
Con never extended jobs turn through Summary.
Per-speaker notes
FeedbackDetailed, specific, calibrated to your paradigm. Tells each debater exactly what to work on.
Alex · Pro Speaker 1
Clean signposting in C — easy to flow. In Summary you under-weighed climate vs. Con's economy frame. Try collapsing earlier.
Signposting +Weighing
Jordan · Con Speaker 2
Strong turn execution. FF was crisp but you dropped extension on mechanism — that lost you the round.
Polished RFD
DecisionA clear winner pick and a paragraph the teams can actually learn from. Ballot-ready.
Decision
Pro wins (Lincoln HS PA)
Reason for Decision
“I voted Pro because the dropped mechanism on the carbon tax meant Con's turn never linked. Pro's climate impact was extended through Summary…”
Why DebateJudge
Built for the way debate actually works.
Real workflow features, not just an AI prompt — every part of judging a round or reviewing a mock round, in one place.
Speech & prep timers built in
No stopwatch juggling. The clock starts and stops with each speech automatically, including prep for both sides.
Refine when the AI is wrong
One sentence to the AI fixes that speech — and improves every remaining speech in the round automatically.
Pause and resume any time
Save mid-round and pick up where you left off later — even from a different device or browser tab.
Lay or tech, any paradigm
Custom round-level instructions ("treat me as lay", "skip greetings") shape every speech the AI flows.
Save what you reuse
Paradigms, team presets, every-round rules — save once, load on every future round. Your judging profile, built up over time.
Individualized feedback
All four debaters get their own paragraph — concrete strengths, specific improvements — not generic copy-paste advice.
Any browser, no install
No signup. No download. No tournament booth setup. Open a tab, start judging, close the tab when you're done.
Your data stays on your device
Audio recordings, evidence files, and your saved rounds live in your browser. We never store them on a DebateJudge server.
Powered by the latest from OpenAI
Cutting-edge AI. Tuned to you.
DebateJudge runs on the newest models from OpenAI — and every output is shaped by your paradigm (or the paradigm you're practicing for). Two people running the same round through it won't get the same flow, feedback, or RFD.
GPT-4o
OpenAI's flagship
The same model behind ChatGPT. Writes your flow per speech, the round summary, per-speaker feedback, and your RFD draft. Reasons over your paradigm to weigh arguments your way.
Whisper
State-of-the-art transcription
OpenAI's speech-to-text. Captures every word of every speech, even fast or accented, so the AI has the full round to work with.
GPT-4o-mini
Fast live-flow model
A leaner, faster variant. Used to flow each speech live as you advance to the next one — so your flow sheet is ready before the next speaker starts.
Adapts to your paradigmEvery output, every round.
Speed & jargon
Parent judge? Plain English, no shorthand. Experienced judge? Full debate vocabulary.
Weighing preferences
If your paradigm prioritizes magnitude over probability, the RFD weighs that way.
Topic familiarity
If you said you're not read up on the topic, every output assumes less context.
Built specifically for debate
Three steps. Every one designed for the way debate actually works.
This isn't a generic transcript app. Every feature, every workflow, every prompt is tuned for PF (and soon LD, Policy, Speech, Congress).
1
Set up the round
~30 sec
Pick PF, enter teams, speakers, and topic. If you need a paradigm, tap ✨ to generate one from 4 quick questions.
- ✓Optional paradigm wizard — designed for parent judges, skippable for everyone else
- ✓Format-aware setup (PF live; LD, Policy, Speech, Congress coming)
- ✓Plain-English protocol notes for every speech
2
Record the round
~35 min
Whether it's a live round or a mock round with your team, DebateJudge listens to every speech and takes the flow for you. Built-in speech timer, prep clocks, and evidence uploads keep the workflow in one place.
- ✓Auto-tracks the full PF speech sequence (8 speeches)
- ✓Prep clocks for both teams + overtime warnings
- ✓Drop in shared evidence — Word, PDF, images
3
Hit Generate
~60 sec
DebateJudge replays the round, writes a clean flow sheet, drafts feedback for every debater, and writes your RFD — all calibrated to your paradigm. You review, set speaker points, and send.
- ✓Per-speech, per-side flow sheet (color-coded)
- ✓Per-speaker feedback with specific improvement notes
- ✓RFD written in your voice, matched to your paradigm
The difference
What changes with DebateJudge.
Applies whether you're judging a live round or reviewing a mock round with your team.
Without DebateJudge
- ✕You take notes the entire round
- ✕Frantic flowing — you write while listening
- ✕30 minutes writing the RFD after
- ✕Generic ballot feedback like “good job”
- ✕You miss arguments while writing
- ✕Evidence scattered across the email chain
- ✕Generic transcript tools that don't know debate rules
With DebateJudge
- ✓DebateJudge takes every note for you
- ✓You just listen and watch the round
- ✓30-second review of an AI-drafted RFD
- ✓Detailed per-speaker, paradigm-aligned notes
- ✓Every word captured — nothing missed
- ✓Evidence vault with side + speech tagging
- ✓Built for debate — knows speech order, evidence rules, paradigms
Help shape this
What's missing? What's wrong?
Every piece of feedback comes directly to me and I read all of it. If something feels off, if the AI got a call wrong, if you have a feature idea, or if you just want to say how it went — I want to hear it. Especially the critical stuff.
Anonymous by default · takes 30 seconds