Anychat

A unified chat interface for multiple AI models powered by Gradio. This application provides access to various leading AI models through a simple tab-based interface.

Getting Started

Prerequisites

• Python 3.8 or higher
• pip package manager

Installation

1. Clone the repository:

git clone https://github.com/AK391/anychat.git

2. Navigate to the project directory:

cd anychat

Running the Application

1. Install dependencies:

pip install -r requirements.txt

2. Start the application:

python app.py

3. Open your web browser and go to:

http://localhost:7860

Features

• Tab-Based Interface: Easy switching between different AI models
• Model Selection: Dropdown menus for selecting specific model versions
• Token Authentication: Secure API key management for each service
• Multimodal Support: Vision capabilities for compatible models

Supported Models

Gemini

• gemini-1.5-pro: Default model for complex reasoning tasks
• gemini-1.5-flash: Fast and versatile performance
• gemini-1.5-flash-8b: Optimized for high-volume, simpler tasks
• gemini-exp-1114: Experimental version with quality improvements

ChatGPT

• GPT-4 Series: Including gpt-4o (most advanced), gpt-4-turbo, and standard gpt-4
• O1 Series: Specialized reasoning models (preview and mini versions)
• Multiple Snapshots: Access to specific model versions with timestamps

Claude

• Claude 3.5: Latest Sonnet (20241022) and Haiku (20241022) versions
• Claude 3: Access to Opus, previous Sonnet, and Haiku versions
• Default set to latest Sonnet for optimal performance

Meta Llama

• Llama 3.2: Models from 1B to 90B parameters, including vision-enabled versions
• Llama 3.1: Models from 8B to 405B parameters
• Vision Support: Multimodal capabilities in selected models
• Requires SambaNova API key from SambaNova Cloud

Grok

• Access to xAI's Grok beta model

Qwen2.5

• 72B parameter model from Alibaba
• Requires Hyperbolic API key from Hyperbolic

Perplexity

Three categories of models:

1. Sonar Online Models (8B-405B parameters)
   • Include real-time search capabilities
   • Requires beta access
2. Sonar Chat Models (8B-70B parameters)
   • Standard chat functionality
3. Open Source Models (8B-70B parameters)
   • Based on Hugging Face implementations

Technical Details

• Built with Gradio Blocks interface
• Supports dynamic model switching
• Implements token-based authentication
• Handles multimodal inputs where supported