AI Voice Analysis

Trust your gut.
Now prove it.

Valid measures the likelihood of deception using real-time AI voice analysis. Baseline calibration. Honest accuracy disclosure. No recordings stored.

Get your validity score See how it works

~80% Accuracy (honestly disclosed)

0 Recordings stored

Real-time Processing only

How Valid works

Baseline. Test. Know.

Three steps. Ten minutes. Clarity you can hold in your hand.

Calibrate the baseline

Ask neutral, verifiable questions — name, birthday, where they grew up. Valid learns what their truth-state voice sounds like. This is the anchor.

→

Ask the real question

Ask what you actually want to know. Valid compares the voice pattern against the calibrated baseline in real time.

→

Receive your validity score

A probability score: how likely is this statement to be truthful? The report validates your intuition — whatever it tells you.

Why Valid

It validates you either way.

If the score is high — you were right to trust.

If the score is low — you were right to doubt.

Either way: you're not crazy. You were paying attention.

That's what "Valid" means.

Validity Score

72%

Statement variance detected

Significant deviation from baseline

◎

No recordings stored Real-time processing only. Data disappears the moment analysis completes.

◎

No two-party consent issues Works on any live conversation — in person or on a call. Nothing is recorded.

◎

Patent-pending methodology Proprietary baseline calibration + AI correlation. Not a generic voice stress app.

◎

~80% accuracy, disclosed honestly We tell you this upfront. It's a probability tool, not a lie detector. Use it as one signal among many.

Reports

Choose your level of certainty.

Same AI. More detail as you go.

Essential

$19.99

A clear, direct verdict on the statement you tested.

Simple result: "More likely deceptive" or "Likely truthful"
Baseline calibration session
Immediate delivery

Get Essential

Detailed

$49.99

Numeric scoring with a breakdown of what the AI measured.

Validity score (0–100%)
Per-phrase deviation markers
Baseline comparison summary
Written narrative interpretation

Get Detailed

Our refund policy says everything.

To receive a refund, simply provide a notarized statement from the person you tested confirming Valid's result was incorrect.

We offer this policy because we stand behind the accuracy of the analysis. We've never had to honor it.

Questions

Everything you're wondering

Yes. Valid processes voice in real time without storing any audio. Because nothing is recorded, two-party consent recording laws don't apply. You're using AI analysis of a live conversation you're a participant in — the same way you'd use your own memory or notes.

~80% accuracy on controlled testing. We disclose this honestly because we believe you deserve to know. Valid is a probability tool — one strong signal among others. It's most useful when it confirms what your gut already suspects.

No. Valid analyzes voice patterns in real time through your device. No app is installed on their side. It's invisible to them.

Before asking the real question, you ask 3–5 neutral questions you already know the answers to: their name, birthday, hometown. This gives the AI a voice fingerprint for their truthful state. The more baseline questions, the more accurate the comparison.

No audio is ever stored. Processing happens in real time and the audio data is discarded immediately. Your report contains only the analytical output — no voice recordings, no audio files.

Provide a notarized statement from the person you tested confirming the result was incorrect. That's our refund condition. It sounds unusual because it is — it reflects our confidence in the analysis.

Trust your gut. Now prove it.