🤖 Auto Chat 🤖

An AI vs. AI conversation simulator built with Tkinter.

Engage two AI personas in turn-based conversations using various Large Language Models (LLMs).

---

🌐 Web UI (recommended)

The modern way to run Auto Chat: a React web interface backed by FastAPI, with real-time streaming over WebSockets.

pip install -r requirements.txt
python web_server.py            # serves the pre-built UI at http://127.0.0.1:8008

• 👥 2–10 personas per conversation — each on its own model, from any provider (mix Ollama, LM Studio, OpenAI, Anthropic, OpenRouter, Venice AI, and Grok/xAI in one chat)
• ⚡ Live token streaming, typing indicators, markdown rendering
• 💭 Visible thinking — reasoning models' chain-of-thought streams live in collapsible blocks under each message
• ▶️ Continue / redo / resume — extend a finished conversation, regenerate the last turn, or reload any past conversation from history and keep it going
• ✨ AI persona generator — describe a character in one line, a model drafts the full persona for the library
• ✦ One-click summaries — a model writes a recap of the conversation so far
• 🎛️ Pause / steer topics / inject narrator events mid-conversation; per-persona temperature & max-tokens; adjustable turn delay
• 📚 Templates, searchable history with favorites, token & cost dashboard
• ⭳ Export transcripts as Markdown / JSON / text

To hack on the frontend: cd web && npm install && npm run dev (Vite dev server with proxy to the backend), and npm run build to refresh web/dist.

The original Tkinter desktop app still works: python auto_chat.py.

---

🌟 Features

• 🗣️ Dual AI Conversations: Pit two distinct AI personas against each other.
• 🔌 Multi-LLM Support: Integrates with Ollama, LM Studio, OpenAI, Anthropic, OpenRouter, Venice AI, and Grok/xAI APIs.
• 🎭 Persona Management: Easily create, edit, and manage AI personalities.
• 🎨 Modern GUI: Built with Tkinter and styled with ttkbootstrap for a clean look.
• 💾 Conversation Logging: Save and export chat transcripts.
• 🎤 Narrator Mode: Interject system messages or context into the conversation.
• 🎯 Topic Control: Guide the conversation with specific topics.
• ⚙️ Persistent Configuration: Saves your settings and API keys.
• 🖥️ CLI Mode: Run conversations from the terminal with cli_chat.py.

---

🛠️ Requirements

• Python 3.8+
• Tkinter & ttk (usually included with Python)
• ttkbootstrap
• requests
• python-dotenv
• rich (for the Persona Generator)

---

🚀 Installation

Click to expand installation steps

1. Clone the repository:

   git clone https://github.com/rustyorb/auto_chat.git
   cd auto_chat

   Or download the source code ZIP.

2. Create and activate a virtual environment:

   # Create virtual environment
   python -m venv venv
   
   # Activate virtual environment
   # On Windows:
   .\venv\Scripts\activate
   # On macOS/Linux:
   source venv/bin/activate

   Note: Using a virtual environment is highly recommended!

3. Install dependencies:

   pip install -r requirements.txt

4. Set up personas.json:

   # Copy the example personas file
   cp personas.json.example personas.json

   (Alternatively, rename personas.json.example to personas.json manually)

5. Set up config.json:

   # Copy the example config file
   cp config.json.example config.json

   (Alternatively, rename config.json.example to config.json manually)

   ✨ The example files provide a great starting point. Edit them directly or use the app's interface!

---

▶️ Usage

💬 Main Application (auto_chat.py)

1. Ensure your virtual environment is activated (source venv/bin/activate or .\venv\Scripts\activate).
2. Launch the application:

   python auto_chat.py

3. Setup Screen:
   • 👤 Select two personas.
   • ☁️ Choose API providers and models for each.
   • 🔢 Set the max number of conversation turns.
   • 📝 Enter a conversation topic.
4. During Conversation:
   • ⏯️ Pause/Resume
   • 📢 Add Narrator Messages
   • ⏹️ Stop Conversation
   • 📄 Save Log

🧑‍🎨 Persona Generator (persona_generator.py)

A handy tool for crafting new AI personalities:

python persona_generator.py