> For the complete documentation index, see [llms.txt](https://mycroft-ai.gitbook.io/docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://mycroft-ai.gitbook.io/docs/mycroft-technologies/mimic-tts.md).

# Mimic TTS

![](/files/c41DVJ1OrGCFhm7BVAfN)

Over the years, Mimic like other Mycroft components has gotten clearer, faster, and more flexible.

You can hear samples from all three Mimic systems below, speaking the same [sentence](https://clagnut.com/blog/2380/#English_phonetic_pangrams).

<table><thead><tr><th width="153">Name</th><th width="150">Sample</th><th width="150">Local</th><th>Technology</th></tr></thead><tbody><tr><td><img src="https://mycroft.ai/wp-content/uploads/2017/02/mimic.png" alt="" data-size="line"> <a href="/pages/-LouLmAKOi_GveuIYirR">Mimic 1</a></td><td><a href="https://mycroft.ai/wp-content/uploads/2022/04/mimic1_pangram.wav">Mimic 1 sample</a></td><td>Yes</td><td><a href="http://www.festvox.org/flite">Festival Lite (flite)</a></td></tr><tr><td><img src="https://mycroft.ai/wp-content/uploads/2019/05/mimic2-logo.svg" alt="" data-size="line"> <a href="/pages/2Go7A2XL7JlTwXJK0ecS">Mimic 2</a></td><td><a href="https://mycroft.ai/wp-content/uploads/2022/04/mimic2_pangram.wav">Mimic 2 sample</a></td><td>No</td><td><a href="https://github.com/keithito/tacotron">Tacotron</a></td></tr><tr><td><img src="https://mycroft.ai/wp-content/uploads/2020/10/Mimic_color.png" alt="Mimic logo" data-size="line"> <a href="/pages/PEM5BLkkLbPx19wTZFhv">Mimic 3</a></td><td><a href="https://mycroft.ai/wp-content/uploads/2022/04/mimic3_pangram.wav">Mimic 3 sample</a></td><td>Yes</td><td><a href="https://github.com/jaywalnut310/vits">VITS: Conditional Variational Autoencoder with Adversarial Learning</a></td></tr></tbody></table>

### COMING SOON: Mimic 3

Mimic 3 is a privacy-focused open-source neural Text to Speech (TTS) engine that can run faster than real-time on low-end devices like the Raspberry Pi 4. In human terms that means it sounds great, it can run completely offline or in the cloud, and you can verifiably trust it.

{% content-ref url="/pages/PEM5BLkkLbPx19wTZFhv" %}
[Mimic 3](/docs/mycroft-technologies/mimic-tts/mimic-3.md)
{% endcontent-ref %}

### Mimic 2

Mimic 2 is our older machine learning TTS engine designed to run in the cloud. It has been the default voice for most installations of Mycroft for many years.

{% content-ref url="/pages/2Go7A2XL7JlTwXJK0ecS" %}
[Mimic 2](/docs/mycroft-technologies/mimic-tts/mimic-2.md)
{% endcontent-ref %}

### Mimic 1

Mimic 1 is a fast, light-weight TTS engine based on Carnegie Mellon University's FLITE software. It concatenates speech sounds to create complete phrases.

{% content-ref url="/pages/-LouLmAKOi\_GveuIYirR" %}
[Mimic 1](/docs/mycroft-technologies/mimic-tts/mimic-overview.md)
{% endcontent-ref %}

### Mimic Recording Studio

Mimic Recording Studio provides simplifies the collection of voice training data from individuals, by allowing you to speak and record phrases from a *corpus*. This data can then be used to train a voice model using Mimic 3, Mimic 2, or another machine learning TTS engine.

{% content-ref url="/pages/-LocrTXhueLUfSmpUHTs" %}
[Mimic Recording Studio](/docs/mycroft-technologies/mimic-tts/mimic-recording-studio.md)
{% endcontent-ref %}