Aktiva AI is a versatile self-hosted Discord bot designed to enable users to interact seamlessly with multiple AI personalities in a single environment.
Talk to multiple AI characters through one bot with unique avatars and personalities for each interaction.
Each channel and thread has its own dedicated memory for an immersive interaction experience.
Integrated with Microsoft's Florence-2 AI for powerful multimodal capabilities and image analysis.
Upload PDF documents for AI characters to read, analyze, and provide insights during interactions.
Powered by DuckDuckGo to perform live web searches and retrieve real-time information.
Add unlimited characters with support for SillyTavern and Pygmalion AI character formats.
The simplest way to get Aktiva AI up and running is using Docker. Just follow these steps:
git clone --branch main https://github.com/Iteranya/AktivaAI.git cd AktivaAI
Create a .env file based on the example.env provided in the repository.
docker compose up -d
In your Discord server, use the command:
/set_whitelist Vida-chan
For those who prefer manual setup, follow these steps:
git clone --branch main https://github.com/Iteranya/AktivaAI.git cd AktivaAI
python3 -m venv venv source venv/bin/activate # On Windows, use `venv\Scripts\activate`
python3 -m pip install -r requirements.txt
Create a .env file based on the example.env provided.
python bot.py
Aktiva AI comes with default characters and supports importing any character in SillyTavern or Pygmalion format. Create your own or import existing ones for endless possibilities.
The default assistant character, helpful and knowledgeable about the bot's features.
Import any character in SillyTavern or Pygmalion format for endless possibilities.
Create your own characters with custom personalities, backstories, and behaviors.