Confidence
An even pace, few fillers, controlled pauses and a voice that does not waver.
Hear how your voice comes across
Free voice analysis for talks, interviews and presentations. Record yourself or drop in a file, and get eight scores on your delivery — confidence, charisma, clarity and more — each with the measurements behind it.
- Private — your audio never leaves this device
- Explainable — every score shows its evidence
- Free — no account, no sign-up
Drop a recording here, or click to choose a file
any audio or video file your browser can play
or record now
0:00
What it measures
Eight reads of your vocal delivery, built from pitch, loudness, pauses, speaking rate, filler words and the cycle-to-cycle steadiness of your voice.
Charisma
Pitch range, moving intonation, dynamic loudness and a resonant voice.
Authority
An unhurried pace, pauses that let points land and clear projection.
Clarity
A clean signal, a moderate rate, distinct articulation and clear phrase breaks.
Warmth
Melodic intonation, a smooth unstrained voice and an unhurried pace.
Enthusiasm
Punchy loudness dynamics, wide animated pitch and a lively pace.
Professionalism
Fluent, low-filler speech, a composed pace and structured pauses.
Nervousness
Micro-tremor, unstable loudness, hesitations and pressured speed. Lower is better.
How it works
- Measure. Your recording is decoded in the browser and measured frame by frame: pitch, loudness, pauses, jitter, shimmer and harmonics-to-noise ratio.
- Score. Each dimension is a weighted average of simple, published curves over those measurements. Nothing is trained or hidden, so every score can be traced back.
- Follow the arc. The same scoring runs again on a sliding window, so you can see whether you settled in or tailed off — not just a single average.
Private by design
Everything runs on your device, in WebAssembly. Your recording is never uploaded, stored or shared — close the tab and it is gone.
Read it honestly
These are indicative reads of delivery, not facts about you. Accent, language, microphone and room all move the numbers, and the bands are set for English speech. The reliable use is your own trend over recordings made the same way.
Getting a useful recording
- Speak for one to three minutes on one subject, as you would to an audience.
- Use a quiet room and keep a steady distance from the microphone.
- Turn off noise suppression and auto-gain if you can — they flatten the loudness dynamics being measured.
- Keep the set-up identical between recordings so the comparison means something.
Frequently asked questions
Is my recording uploaded anywhere?
No. Decoding, measurement and scoring all run inside your browser. If you tick transcription, the speech model is downloaded to your browser once; your audio still stays on your device.
How long should my recording be?
One to three minutes of continuous speaking on one subject works best. Clips shorter than about 14 seconds get scores but no over-time view.
What do the scores actually measure?
Vocal delivery only: pitch, loudness, pauses, speaking rate, filler words and voice steadiness. Each of the eight scores is a transparent weighted combination of those measurements, and the page shows which ones moved it. It does not judge what you say.
Can I compare my score with someone else's?
Not meaningfully. Accent, language, microphone and room all move the numbers. The reliable use is tracking your own trend across recordings made under the same conditions.
Which file formats work?
Anything your browser can decode: WAV, MP3, M4A, AAC, FLAC, OGG, Opus and WebM, and the audio track of MP4 and MOV videos. If a file is refused, convert it to WAV.
Does it cost anything?
No. It is free and needs no account.