Speech Lab

Speech Lab allows to clone voice with 10 seconds audio and use it in videos created by Astrid AI bots

What is Speech Lab?

Speech Lab is a set of tools within Astrid AI, which allows to clone any voice with 10-seconds long audio and use it in your creations through Text-to-Speech and Speech-to-Speech technologies.

Voice Cloning

You can clone a voice by providing short audio record. Although it is recommended to provide 1-2 minutes long clear voice recording, you can have similar results with short audio as well.

Requirements

1. Audio with the voice in the following formats: audio, video, audio-message

2. Voice should be clear, without background noices.

3. There should only one person speaking.

4. Provided media should comply with Voice Lab Terms of Use

Step-by-Step instruction

1. Open Astrid AI bot and press /start

2. Open Speech Lab 🎙 panel through the button below the start message. If you have enough credits, press ➕ New Voice

3. Send Voice Sample in supported format.

4. Give a tag to this voice. It can be the name of the person or anything that helps you identify it.

5. Press ✅ Clone voice button

6. Done! Now you can use this voice in any of the videos you create.

Using the cloned voice

Voice Changer

You can change the voice on the video without changing the speech itself. To do this, simply choose Speech Lab 🎙 tab before submitting the job and choose one of the cloned voices.

Text-To-Speech

Text to speech allows to change what person on the video is saying by providing prompt text. To do that, simply send a text after choosing the desired voice.

Speech-to-Speech

Speech-to-Speech is the most advanced AI speech technology, because it has what Text-to-Speech lacks - Emotions. With Text-to-Speech, you can't provide all the emotions of the person and other sounds human can make, but with STS, it is now possible.

It works pretty much the same as TTS, but instead of the text you need to send an audio. It is important to make sure the duration of the speech will not be longer than the video itself.

Results

Bot will replace original audio of the video with the new one generated by AI. After that, you can use advance editting tools to add background sound and others.

Please note, that output is provided by third party service, Elevenlabs, and we're not able to predict the output.

Availability

Voice cloning feature is only provided to users with subscription. Cloned voices are available to all subscribers as long as they are using the bot. Due to limited quota on cloned voices, we are not able to store cloned voices forever, so if you stop using the bot, we retain the right to remove your cloned voices and provide access to other premium users.

Pricing

As of May 2024, the price to clone one voice costs from 1200 to 1500 credits, one time purchase. It is recommended to purchase Premium Subscription which gives 1000 credits daily to be able to clone voices for the cheapest price.