Redtok

Note: This repository is a modified and enhanced version of NarReddit.

Redtok is a script that automates the process of converting top-rated Reddit posts into engaging, Subway Surfers-style TikTok narrations. It supports running as a service and creating multiple videos a day and upload them to a Google Drive folder for instant access and posting on social media.

Features

• Content Generation: Redtok uses GPT-4 to enhance and translate Reddit posts, enabling the creation of multilingual content.

• Reddit Scraping: The script scrapes your chosen subreddit to find the highest-rated post.

• Intelligent Text-to-Speech (TTS) Audio Generation: Redtok generates TTS audio from the post, intelligently determining the speaker's gender and adjusting the TTS voice accordingly.

• Subtitles Creation: Redtok can also create subtitles for the generated audio.

• Video Overlays: The script overlays the generated audio and subtitles onto a selected background video, creating a complete, ready-to-upload video.

Setup

Follow these steps to set up:

1. Install ffmpeg and make it accessible via the $PATH environment variable. More information

2. Install the required dependencies by running pip install -r requirements.txt.

3. Place your preferred video files to be used as backgrounds in the background-videos directory.

4. Create a new Reddit application at this link. Choose 'script' from the radio button menu and set the redirect URI to http://localhost:8080.

5. Set up the Gentle forced aligner and start the server.

6. Sign up for an Elevenlabs account if you don't have one already.

7. Obtain an OpenAI API key.

8. Create a .env file using the .env.template provided in the repository.

9. Populate the .env variables as per your requirements.

   • Note: You can adjust the character limit for posts in the .env file. It's advisable to use a limit other than the defaults to avoid exhausting your Elevenlabs API character limit too quickly.

   • Note: You'll need an API key with access to GPT-4 for this feature. Bear in mind that using GPT-4 may result in higher API costs.

10. Run the main script!

• Note: This current version is setup to upload created videos directly to a Google Drive folder, which needs to be set in the .env file, and will require you to create a Google Cloud app. This can be removed from the main script if you only want to store the videos locally.

• Note: The main script is meant to be ran as a background service, and will generate one video per specified subreddit in the .env file per day at 1pm (this can be changed inside app.py). Run the script when evertything is ready!