Skip to generator

Help, without the guesswork

AI Voice Generator FAQ

From your first recording to your next download. Find the details that help you create with confidence.

Getting started

Choose the right tool for your next idea.

Which voice tool should I use?

Use text to speech to read a script with a preset voice. Use voice cloning to read new words in your own or an authorized voice, starting with a short reference recording. Use the voice converter to change the voice in an existing recording while keeping its delivery. Compare the three tools.

Can I try it without an account?

Yes. Guests have a shared daily allowance of 500 characters for text to speech and voice cloning, plus 30 seconds of voice conversion. A free account receives 1,000 daily credits and a one-time 1,000 sign-up bonus. Usage limits still apply; free audio is for personal, non-commercial use. See free allowances and paid plans.

Which languages are supported?

The current public tools support English. Start with a short English script, then listen for names, pronunciation and timing before making a longer version. We do not promise reliable results in other languages. Try English text to speech.

Does the voice converter work during live calls?

No. It processes an uploaded or recorded clip, rather than changing a live call or game microphone. The current source recording limit is 30 seconds and 2 MB per request. Open the recorded voice converter.

Recording & results

A clear recording gives your voice a better starting point.

What makes a good voice-cloning sample?

Record or upload 5–20 seconds of one person speaking naturally; about 10 seconds works well. Use a quiet room, keep the microphone at a steady distance and avoid music, overlapping speakers and heavy effects. Only use your own voice or one you have permission to clone. Record a voice reference.

Do I need to type the reference recording word for word?

The cloning tool can recognize the reference speech automatically. Check and correct the recognized text if needed so it matches what is actually said. Your new script belongs in the separate text field; it does not replace the reference transcript. Open voice cloning.

How long can my script be, and can I download the result?

Text-to-speech and cloning requests currently accept up to 500 characters each. You can play the completed result and download an MP3. Download important audio promptly: result links expire, and account history is not permanent storage. Read about result storage.

Why does a voice sound different from the sample?

AI speech can vary with the reference recording, wording and punctuation. It may mispronounce a name, miss an expression or sound different from the original speaker. Try cleaner speech and a short, simply punctuated script. Listen to the whole result before publishing it; a perfect match is not guaranteed. Get help with a generation.

Credits & payments

Know what you are using before you generate or buy.

How are credits used?

Standard text-to-speech and cloning cost 1 credit per character. Voice conversion uses 1,000 credits per minute, based on the recording length. The tool shows the estimate before generation. Paid pack validity, monthly allocations and rollover rules are listed with each plan. Compare credits and plans.

Can I use the audio commercially?

Free audio is for personal, non-commercial use. Commercial use is included on all paid plans, including Starter, subject to the Terms and your rights to the content and voice. Read the usage terms.

What happens if a generation fails?

When a generation fails, its reserved credits or allowance are returned. If your balance is not restored, contact support with the generation reference and approximate time. A successful result that you do not like is different from a technical failure and does not automatically qualify for a refund. Contact support.

How do cancellation and refunds work?

Cancel a subscription through the billing portal in your account to stop future renewals. Annual plans allocate credits monthly. To request a refund, contact support within 24 hours of payment while all paid credits covered by that payment remain unused. Cancellation does not itself issue a refund; requests are reviewed under the refund policy and applicable consumer rights. Read the full refund policy.

Privacy & permission

Be clear about whose voice it is and how it is handled.

Whose voice can I clone or convert?

Use your own voice or get permission from the voice owner. For conversion, permission must cover both the source recording and the target voice. Do not impersonate someone deceptively, commit fraud or infringe their rights. A public demo is not a license to reuse that person's voice. Read the acceptable use rules.

Are my recordings sent to other providers?

Yes. Audio and text are sent to AI processing providers to produce results: SiliconFlow for text to speech and voice cloning, and fal.ai for voice conversion. We do not sell personal data or use voice samples to train our own models. Review the privacy policy before uploading sensitive material. Read how data is processed.

How long can I access my generated audio?

For signed-in accounts, access through generation history ends 30 days after creation; individual signed download links may expire sooner. Access expiry does not necessarily mean immediate deletion from storage. Download anything you want to keep. You can contact support about privacy and deletion requests. Read the retention and deletion policy.

Can I turn off analytics cookies?

Yes. Use Cookie settings in the footer to change your choice. Analytics cookies remain off until you allow them in the EEA and UK; elsewhere they are enabled by default. Rejecting analytics cookies does not block the voice tools. Limited measurement without analytics cookies may still occur, as described in the privacy policy. Read about analytics choices.

A direct line

Some questions need a human.

For account, generation, payment or privacy help, tell us what happened. Please leave out passwords and card details.

Contact support