For AI agents: a documentation index is available at the root level at /llms.txt. Append /llms.txt to any URL for a page-level index, or .md for the markdown version of any page.
Learn how to clone your voice instantly using our best-in-class models.
Creating an Instant Voice Clone
When cloning a voice, it’s important to consider what the AI has been trained on: which languages and what type of dataset. Read more about each individual model and their strengths on the Models page.
Guide
If you are unsure about what is permissible from a legal standpoint, please consult the Terms of
Service and our AI Safety
information for more information.
Under the Voices section in the dashboard, click the My Voices then click Use voice to begin using it.
Best practices
Record at least 1 minute of audio
Record at least 1 minute of audio
Avoid recording more than 3 minutes, this will yield little improvement and can, in some cases, even be detrimental to the clone.
How the audio was recorded is more important than the total length (total runtime) of the samples. The number of samples you use doesn’t matter; it is the total combined length (total runtime) that is the important part.
Approximately 1-2 minutes of clear audio without any reverb, artifacts, or background noise of any kind is recommended. When we speak of “audio or recording quality,” we do not mean the codec, such as MP3 or WAV; we mean how the audio was captured. However, regarding audio codecs, using MP3 at 128 kbps and above is advised. Higher bitrates don’t have a significant impact on the quality of the clone.
Keep the audio consistent
Keep the audio consistent
The AI will attempt to mimic everything it hears in the audio. This includes the speed of the person talking, the inflections, the accent, tonality, breathing pattern and strength, as well as noise and mouth clicks. Even noise and artefacts which can confuse it are factored in.
Ensure that the voice maintains a consistent tone throughout, with a consistent performance. Also, make sure that the audio quality of the voice remains consistent across all the samples. Even if you only use a single sample, ensure that it remains consistent throughout the full sample. Feeding the AI audio that is very dynamic, meaning wide fluctuations in pitch and volume, will yield less predictable results.
Replicate your performance
Replicate your performance
Another important thing to keep in mind is that the AI will try to replicate the performance of the voice you provide. If you talk in a slow, monotone voice without much emotion, that is what the AI will mimic. On the other hand, if you talk quickly with much emotion, that is what the AI will try to replicate.
It is crucial that the voice remains consistent throughout all the samples, not only in tone but also in performance. If there is too much variance, it might confuse the AI, leading to more varied output between generations.
Find a good balance for the volume
Find a good balance for the volume
Find a good balance for the volume so the audio is neither too quiet nor too loud. The ideal would be between -23 dB and -18 dB RMS with a true peak of -3 dB.
FAQ
What is the difference between Instant Voice Cloning (IVC) and Professional Voice Cloning (PVC)?
Professional Voice Cloning (PVC), unlike Instant Voice Cloning (IVC) which lets you quickly clone voices with less than 2 minutes of audio, allows you to train a more realistic model of your voice. This is achieved by training a dedicated model on a large set of voice data to produce a model that’s virtually indistinguishable from your original voice.
Since Professional Voice Clones require fine-tuning and training, it will take some time before you can use your voice clone. Giving an estimate is challenging as it depends on the number of people in the queue before you and a few other factors, but usually fine-tuning will take 3-6 hours.
You will receive an email notification once your Professional Voice Clone is ready.
How many voice samples should I upload for Instant Voice Cloning?
Cloning with instant voice cloning can be a bit complicated, and we do have some general guidelines. However, they are just that: guidelines. We don’t have any set rules when it comes to number of samples or length. We’ve seen users use samples of only 30 seconds and get excellent results, while we’ve also seen some users use 10 minutes of audio and have worse results. But we do have a few things that you should consider.
Audio quality is the most important aspect to consider when using instant voice cloning.
The number of samples is irrelevant; what’s important is the total run time. Having more than 2-3 minutes of audio will yield little improvement and can, in some cases, even be detrimental to the stability of the clone.
What files do you accept for voice cloning?
Recommended:
MP3 192kbps+
Length:
1-2 minutes of good audio for Instant Voice Cloning
30min - 180min of good audio for Professional Voice Cloning
For both Instant Voice Cloning and Professional Voice Cloning, we accept a plethora of file types, but we strongly recommend using MP3 with a bitrate of 192kbps or above. Using an uncompressed format such as WAV will yield little to no improvement. It is instead recommended to focus on the quality of the actual recording to ensure it is recorded professionally without any background noise, room reverb, multiple speakers, at a consistent volume with a consistent tone, no extremely long gaps of silence, and so on.
Are there any restrictions on what voices I can upload for voice cloning?
At Eleven, we’re fully committed both to respecting intellectual property rights and to implementing safeguards against the potential misuse of our technology:
We only partner with clients who adhere to our Terms of Service and Prohibited Use Policy which prohibit malicious use of our technology towards any purpose which can be deemed illegal or harmful;
We seek to support voice owners and their licensors in claiming their rights and all known infringements will be reviewed and actioned;
All audio generated by our models can be instantly traced back to the user responsible for the generation.
The technology we’re developing is new and clear regulation is yet to be introduced. Part of our goal as an AI research lab is to spread awareness about the existence of this technology, its potential, as well as its limitations.
Are there any tips to get good-quality cloned voices?
The bottom line is: good consistent input = good consistent output.
Length
Instant Voice Cloning: 1 - 2 minutes of good audio
Professional Voice Cloning: 30 - 180 minutes of good audio
Use the best and clearest audio clips that you can find. There should only be one speaker without background noise of interference and their voice should be loud and clear.
Instead of using many clips of different quality just to increase the length, prioritize clips where the microphone quality is obviously very high and where the quality and tone is consistent throughout, rather than focusing on increasing the total runtime.
Ensure that most of the dialogue in your clips aligns with the speaker’s speaking style and intonation that you prefer the most. You don’t want too many chunks of dialogue where the speaker deviates from the desired speech patterns you want to hear.
If necessary, use a noise remover to reduce any background noise.
You can find more information in our documentation here.
Can I clone my voice in a language other than English?
Yes, you can clone your voice speaking any language that is supported by the Flash v2.5 and Turbo v2.5 model. You can find the full list of languages here.
You can even clone a voice speaking a language that the AI is not compatible with, but the results might be very unpredictable, as the AI has never heard that language before. However, it will try its best to clone the voice tonality of the speaker, but it will not be able to speak that language. We would not recommend doing this.
Can I export my voice clones?
No, you cannot export your voice clones, and they are only usable on ElevenLabs and not anywhere else.
If you want to save the voice clone to be able to clone it again later, you will have to save the samples that you used to create the cloned voice. Please be aware that each clone will be slightly different, even if the same audio is used.
Why does my voice or accent not sound correct after cloning?
Currently, we offer two choices for cloning:
Instant Voice Cloning (IVC): IVC is less resource-intensive and provides instant
results that you can use immediately. This method is swift, requiring only about 1 to 3 minutes
of audio input for a high-quality clone, and is often ideal for most general uses but might have
trouble with unique voices or accents.
Professional Voice Cloning (PVC): PVC demands significantly more resources and
you are required to provide the AI with a substantial amount of data (between a minimum of 30
minutes and closer to 3 hours for optimal results). This process involves fine-tuning the model
using the provided dataset to create a customized model. The estimated training time is roughly
2-6 hours, but the process may take longer depending on how many other voices are queued for
fine-tuning.
If you have a rather unique voice with a less common accent, instant voice cloning might not provide a perfect replication of your voice. Then the only way to achieve something like that might be through professional voice cloning. Instant voice cloning is generally very accurate, but under certain circumstances, such as those mentioned above, you might have to resort to professional voice cloning to obtain the most perfect clone.
Unfortunately, there is no way to influence the accent or tone of the clone after the clone has already been created; the only way to influence it is to change the actual samples you use for cloning. Just small changes to the samples can make a big difference.
What does the error 'No model found for this voice. Please select another voice' mean?
You will see this error if you try to use your Professional Voice Clone (PVC) before it has completed the fine-tuning process and is available for use.
When you create a PVC, it needs to go through a number of processes before it becomes available for use. After you have verified your voice, it will be processed and queued for fine-tuning. Depending on how many other voices are currently queued for fine-tuning, we estimate that this process will usually take between 3-6, but it can take up to 24 hours.
You can check the progress of your PVC in My Voices by finding the voice in your list of voices, then clicking View to see more details. You can hover over each model to see the current status.
