Creating and Using Custom Characters

Access the limitless potential of your own characters using our simple custom character process

Much of the information on this page is sourced from the DeepFakeAI blog. To read more of our posts, visit blog.fakeai.io

Note: Professional Voice Cloning is not yet released, this page will be updated when the feature is ready for public use

In this guide, we’ll explore how to use DeepFakeAI’s special feature to create custom deepfake characters. You’ll learn to prepare your media files, access the platform, and manage the creation process from start to finish.

Whether for creative projects, marketing campaigns, or personal exploration, this guide will help you navigate the process with ease.

Step 1: Prepare Your Media Files

Before diving into the DeepFakeAI platform, you’ll need to gather two essential files:

Voice Audio: You need to upload audio of your character to train the AI model. This can be done using either Instant Voice Cloning or Professional Voice Cloning. The differences are as follows:

Instant Voice Cloning	Professional Voice Cloning
Requires approximately 8 seconds of audio	Requires around 1 hour of audio
No training required	Takes longer to train, but creates a professional-level voice model
Suitable for short, entertainment clips or circumstances where collecting lots of voice data is impractical / undesirable	Suitable for more convincing and longer clips which require a more detailed vocal synthesis where voice data is readily available

Once you've chosen your cloning model, you must collect clear voice audio in MP3 or WAV format. The audio must be distinct, with minimal background noise, to ensure the AI can accurately train on the character’s vocal traits. This dataset is crucial for the AI to learn various voice characteristics and accents, resulting in a highly accurate voice clone.

Video: Prepare a short video clip (MP4 format, max 2 minutes, less than 10MB) of the character you want to create. This video will be used to generate the visual aspect of your deepfake character.

Tip: The quality of these files directly impacts the accuracy and realism of your deepfake character, so take the time to ensure they are as high-quality as possible.

Step 2: Access the DeepFakeAI Platform

Once your files are ready, visit the DeepFakeAI platform at app.fakeai.io.

Log In: Navigate to the “Create” section of the website. You can log in using various methods, including email, social media accounts, and traditional username/password options. These options ensure easy and secure access for everyone. If you prefer, you can also log in using your crypto wallet on the Ethereum (ETH) blockchain for seamless access. Create Your Character: After logging in, you’ll be prompted to enter a name for your new deepfake character. This name will be visible on your character list once the creation process begins.

Step 3: Payment and Upload

Now that your character is named, it’s time to proceed with the payment and upload your files:

Payment: Creating a custom deepfake character using the Zero-Shot Model is free, however due to the resources required for the Professional Voice option, Professional Cloning costs $25. You can pay using your cryptocurrency wallet. The price may vary, but it is designed to be affordable and accessible to all users.

Upload Files: Upload your audio file, video, and a detailed description of your character. This description is essential as it defines the character’s personality, tone, and nature. The platform also allows you to choose whether your character will be public or private.

Note: Public characters can be explored and used by the broader DeepFakeAI community, adding a collaborative element to your creation.

Step 4: Processing and Verification

After uploading your files and submitting your character description:

Processing: Click ‘Start Processing’ to initiate the creation of your deepfake character. The AI will begin training using the provided data, which may take a few hours depending on the complexity and quality of the inputs. If you selected the "Instant voice cloning", you will be able to instantly access your character.

Email Verification: If you selected professional voice cloning, you will be prompted to enter your email address, where you will receive a notification once your character’s training is complete. Keep an eye on your inbox for a confirmation email from DeepFakeAI, letting you know when your character is ready for use.

Step 5: Interacting with Your DeepFake Character

Once your character is ready, the fun begins:

Interaction: You can interact with your deepfake character using commands via the web app or Telegram.

In the web app, simply navigate to your character and type out your text to generate responses. On Telegram, you can use the /say command to make your character speak a specific line, or /ask to have it answer questions. Access: To interact with your character, you must either hold $FakeAI tokens or purchase minutes for use on the platform. The cost is approximately $2 per minute of interaction, offering flexible options based on your needs.

Fine-Tuning: If your character has professional voice cloning and if you’re not entirely satisfied with your character, DeepFakeAI offers a fine-tuning option. You can return to the platform to adjust the voice settings, upload new videos, edit the character description, and retrain the character. This retraining process currently costs $25.

And that’s it! You’re now ready to start creating your custom deepfake characters with DeepFakeAI. From preparing your media files to interacting with your unique creations, you have the tools and steps needed to bring your ideas to life. Enjoy exploring the possibilities and making the most of DeepFakeAI’s features!