How to clone a voice without paying a monthly fee

Why the usual way fails

Most voice cloning services bundle a set number of characters or minutes into a monthly tier, and unused allowance usually disappears at the end of the month. That model suits studios that publish every day, but it punishes people whose work comes in bursts. Pricing pages often mix characters, minutes and seats, which makes it hard to compare tools or estimate a single project. Running an open source model yourself avoids the fee but means installing software and keeping a graphics card busy, which is more than most narrators want to manage. So people either overpay for a plan they barely touch or give up on a cloned voice for small projects altogether.

How Resona helps

Resona runs in the browser and works on prepaid credits instead of a plan. You buy a credit pack, the balance never expires, and each feature takes credits only when you use it. Cloning a voice costs a fixed amount once, and after that the voice can be reused for new text at the normal speech rate. If a generation fails, the reserved credits go back to your balance and the attempt stays visible in your usage log. Text to speech is billed by the size of the text you send, so a short script costs little and a long book costs more, with nothing charged in the months you are not working.

Step by step

1. Sign up with email and buy the smallest credit pack to start.
2. Record clean speech from one speaker, ideally one to three minutes with no music or echo.
3. Upload the sample in the console to create a reusable cloned voice.
4. Type or paste your script and generate audio with the new voice.
5. Listen to the result, fix any words in the script that came out wrong, and generate again.
6. Download the file and check the usage log to see what each step cost.

Tips

• Spend your effort on the recording, not on retries. A short clean sample gives a recognisable clone, and a longer one gives a closer match, so record a little more than the minimum.
• Record in a quiet, dry room. Background music, room echo, a second voice or a clipped level all end up baked into the voice model.
• Estimate a project before you buy. The pricing page lists typical costs per minute of narration, which helps you pick a pack size instead of guessing.
• Only clone a voice you own or have written permission to use, and label synthetic speech where a platform asks for it. Keeping that consent in writing protects you if a client or platform asks later.

Questions