Documentation

Create an account, learn the free plan, make your first clip, and find it in your library. No credit card, no setup.

The editor, choosing an engine by character, directing a read with bracketed cues, the fine controls, and Enhance.

Clone your voice from 10 to 20 seconds of clean audio. What to record, the consent attestation, language support, and what delete actually does.

Upload or record a take, pick a clone you built, and convert. What carries over, the real caps, and how usage meters by audio length.

Transcribe a recording, translate the script, and re-voice it in one of 8 languages. Three real steps with an editable transcript between each.

Import a manuscript, split it into chapters and segments, render each line as real audio, and export stitched chapter WAVs. Beta, documented honestly.

Upload or record audio and get an accurate plain-text transcript from a Whisper-class model. Formats, the 25 MB cap, and what metering really counts.

Describe a mood and get a finished instrumental clip from Lyria 3 (preview). Prompt tips, what the tool makes today, and what it honestly does not.

Install, pair with a one-time code, hold a key, and clean text lands at your cursor in any app. How metering, privacy, and the beta caveats actually work.

From download to your first placed sentence: the SmartScreen prompt, the pairing code, the hotkey, and what each part of the little pill means.

The known dead ends and their fixes: SmartScreen, busy microphones, stolen hotkeys, apps that fight paste, Signed out, and the allowance meter.

The full roster by character, plus how to read the benchmark: third-party quality scores and latencies we measure ourselves.

What a character buys, when the meter resets, what happens at the limit, and the real caps on uploads and scripts.

You own your generations: commercial use, worldwide, no watermark, no attribution. What we store, what delete does, and your cloning responsibilities.

Every format each tool reads and writes, with the real caps: 25 MB uploads, 20 MB clone clips, 150,000-character manuscripts, MP3 and WAV out.

The three plans and their real allowances, what counts toward the meter, how checkout and cancellation work, and what happens when you hit the cap.

Share your own clips with explicit consent, remix anyone's script into your studio, join the discussions, and see exactly how moderation works.

No child stock voices, no cloning minors, no child-like presets. The five reasons: consent, privacy law, voiceprint law, replica law, and the abuse record.

Why consent is the legal foundation of voice cloning: publicity rights, voiceprint statutes with real damages, GDPR biometrics, and what genuine permission looks like.

Famous voices are the most legally protected voices, not the least. The ELVIS Act, the FCC's robocall ruling, the New Hampshire deepfake fine, and where parody actually stands.

Death does not end a voice's legal protection. Postmortem publicity rights, California's digital replica law, how estates actually license voices, and what we allow.

The labeling rules arriving for AI audio: the EU AI Act's transparency article, US state election-deepfake laws, platform policies, and where Cantari already stands.

The family-emergency voice scam, the case that reached Congress, what the FTC and FBI advise, the passphrase habit worth adopting today, and what we do on our side.

The best case for voice cloning: people banking their voices ahead of ALS, performers licensing replicas on their own terms, and how you will preserve a voice here when cloning ships (coming soon).

Text to speech (TTS) is software that turns written words into spoken audio. What modern neural engines actually do, and how to size a script in minutes.

Speech to text (STT) turns recorded speech into written text. How modern transcription models work, what limits their accuracy, and what they cannot do yet.

A cue is a stage direction in square brackets, like [whispering], that tells a voice engine how to deliver the next line. Which engines act them, and which ignore them.

Normalization rewrites numbers, dates, and abbreviations into the words a voice should actually say: Dr. into Doctor, 1982 into nineteen eighty-two. Why it matters for long-form audio.

Voice drift is the slow change in a synthetic narrator's tone, pacing, or energy across long-form audio. Why it happens, and how chaptered workflows keep hour seven sounding like hour one.

TTS latency is how long an engine takes to speak. The number depends entirely on where you stop the clock: the first streamed byte, or the complete audio file.

Quality Elo is a listener-vote rating for voice engines, borrowed from chess: blind pairwise comparisons produce a score nobody can self-award. How to read one, with the attribution.

Voice cloning consent is the speaker's explicit permission before their voice is cloned. What good consent practice looks like, and how it is enforced here.

Dubbing replaces the voice track in a new language; subtitling keeps the original audio and translates on screen. The honest trade-offs, and which one this studio does.

Zero data retention means an AI provider processes your request and keeps nothing: no stored prompts, no stored outputs, no training on your text. What ZDR covers, and what it does not.