Record or upload a clear sample. Give your private voice a name, then use it in the voice editor.
Upload or record your voice
30–60 seconds. One speaker, a quiet room, no background music.
MP3, WAV, or M4A · 32 KB to 10 MB
Name your voice
Choose the language spoken in your recording.
Checking account…
YOUR SOUND, ON YOUR TERMS.LOVO AI
LOVO AI
Keep a narrator, not a finished sentence.
A private voice lets you produce a new introduction, corrected instruction, or updated explanation without reusing the words in the reference recording. LOVO AI keeps that voice in your account. After creation, test it on this page with a short script before preparing the rest of your narration.
30–60 s
A clear sample. One speaker. A quiet room.
MP3 / WAV / M4A
32 KB – 10 MB
Private to your account. Subscription required.
01 / MP3 · WAV · M4A
Record the voice you want to reuse
Choose a quiet space and a comfortable speaking style you can sustain. Record 30–60 seconds of one person, or select an MP3, WAV, or M4A file between 32 KB and 10 MB. Listen through the sample first. Music, another speaker, and a distant microphone can all become part of the reference you give the model.
Keep the delivery representative of your project. A calm lesson sample is a better starting point for a calm lesson than a dramatic performance. You must have the speaker's permission and the necessary rights to the recording. Check that permission covers the intended scripts before submitting the file.
Open the project menu and choose Export. Check the preview before saving the file to your device.
Try a video introduction
02 / LOVO AI
Test the awkward words first
Name the voice so you can identify it later and choose the language in the sample. Once creation finishes, the voice is selected in the editor below the workbench heading. Try a sentence containing the names, dates, or technical terms that matter to your project. Check the credit quote before you generate.
Listen for intelligibility and a steady reading rather than a perfect imitation. Rewrite a difficult abbreviation, separate a long instruction, or adjust punctuation before another take. There is no automatic translation: write the text in the language you want spoken and review whether the result suits the speaker and audience.
Workflow illustration · audio is added to the video editor manually
03 / MP3 → WebM
Put the approved take beside your footage
Save the finished MP3. For a narrated video, open LOVO's browser video editor and add that audio file with your local video. Preview the relative levels before exporting WebM. Speech generation processes your text through the voice service; video editing uses the files on your device. The two stages have different data flows.
Narration starts at the beginning of the video, and the export ends with the video. Prepare any required offsets or cuts in your files first. Keep your script and audio after export, so the next revision starts from an approved version. A private voice remains useful across revisions while your subscription is active.
A silent tour of the editor, voices and language options. Voice cloning requires an active subscription.
Use cases
Design the revision before you record
These example briefs show where a reusable narrator can help. They are not claims about customer results or a promise of matching a human performance.
01
VIDEO & SOCIAL
Correct one instruction
A walkthrough has a new menu name. Write the corrected passage, check the product vocabulary, and make a replacement audio file with your private voice. Rebuild the relevant video source as needed; LOVO does not splice individual words into an old recording.
YOUR WORDS
Open the project menu and choose Export. Check the preview before saving the file to your device.
02
COURSES & TRAINING
Welcome the next class
A course needs a new opening for the next group of learners. Keep the wording specific to this lesson and use a voice whose speaker approved that use. Check the full introduction for pronunciation before adding it to the course video.
YOUR WORDS
Today we will work through one practical example. Keep your notes nearby and pause the lesson whenever you want to try a step.
03
PRODUCTS & SMALL TEAMS
Explain the next release
A small team updates its product demonstration. Plan a short narration around the actual screens, then compare the generated take with the recording. Review pacing and factual accuracy together; a familiar voice cannot correct an outdated instruction.
YOUR WORDS
This update adds a simpler way to organize your projects. Let us walk through the new view and the settings you can change.
BEFORE YOU PRESS RECORD
Samples, access, and the next revision
Check the recording and account limits before creating a voice you plan to use again.
An active Starter, Studio, or Enterprise plan unlocks private voice creation and use. Saving a voice occupies a slot; making a new voice also counts against the separate UTC-month creation allowance. Generating narration then uses speech credits. The table below lists both limits, so compare the voices you need to keep with how often you expect to create a new one.
Voice cloning allowances
Plan
Voice slots
Creations per UTC month
Starter
1
2
Studio
3
6
Enterprise
10
20
Can I fix a poor reference after creating the voice?+
Listen before submitting, because a noisy reference is not repaired by writing a better script. If a created voice does not meet your needs, check your remaining creation allowance before making a new one. Removing a saved voice frees its slot but does not return a creation already used that month. Failed attempts report whether the allowance was restored.
Where is the sample sent and who can find the voice?+
LOVO forwards the reference audio to its speech provider, ElevenLabs, for processing and does not retain the original sample itself. The resulting voice belongs to your account library and is not published in the public catalog. Generated scripts and completed audio are kept in your generation history. The Privacy Policy explains processing and retention; private access does not mean that generation runs on your device.
What happens when my plan ends?+
Your saved voices remain associated with your account, but creating or generating with a clone requires an active subscription. Purchased speech credits alone do not unlock that access. Check the current plan and available models before the next revision. Download finished narration you need to keep, rather than relying on an active subscription as the only copy of a project.
Can I publish a take without reviewing it?+
Always listen first. A synthetic take can misread a name, shift an emphasis, or sound different from the reference. Check meaning, pronunciation, and permission for the intended use before publishing. Do not use a cloned voice to impersonate a person or mislead listeners. Some samples need verification the current flow cannot complete; a successful upload does not guarantee creation.
