AI Assist

A Rust library for interacting with multiple AI providers including Google's Gemini API, OpenAI's GPT, and Anthropic's Claude. This library provides a simple and unified way to integrate AI capabilities into your Rust applications.

Features

• Support for multiple AI providers:
  • Google Gemini
  • OpenAI GPT
  • Anthropic Claude
• Simple and unified API
• Async support
• Error handling with anyhow
• Type-safe request and response handling
• Configurable timeouts

Installation

Add this to your Cargo.toml:

[dependencies]
aiassist = "0.1.0"

Usage

First, you'll need to get API keys from the providers you want to use. Then you can use the library like this:

use aiassist::{GeminiClient, OpenAIClient, ClaudeClient, ClientConfig, AIClient};

#[tokio::main]
async fn main() -> anyhow::Result<()> {
    // Configure clients
    let gemini_config = ClientConfig {
        api_key: std::env::var("GEMINI_API_KEY")?,
        timeout_secs: 30,
    };
    
    let openai_config = ClientConfig {
        api_key: std::env::var("OPENAI_API_KEY")?,
        timeout_secs: 30,
    };
    
    let claude_config = ClientConfig {
        api_key: std::env::var("ANTHROPIC_API_KEY")?,
        timeout_secs: 30,
    };

    // Initialize clients
    let gemini = GeminiClient::new(gemini_config);
    let openai = OpenAIClient::new(openai_config);
    let claude = ClaudeClient::new(claude_config);

    // Generate content using different providers
    let prompt = "Tell me about Rust programming language";
    
    let gemini_response = gemini.generate_content(prompt).await?;
    println!("Gemini: {}", gemini_response);
    
    let openai_response = openai.generate_content(prompt).await?;
    println!("OpenAI: {}", openai_response);
    
    let claude_response = claude.generate_content(prompt).await?;
    println!("Claude: {}", claude_response);

    Ok(())
}

Environment Variables

• GEMINI_API_KEY: Your Google Gemini API key
• OPENAI_API_KEY: Your OpenAI API key
• ANTHROPIC_API_KEY: Your Anthropic API key

Provider-Specific Details

Google Gemini

• Uses the Gemini Pro model
• Requires a Google AI Studio API key

OpenAI

• Uses the GPT-3.5-turbo model by default
• Requires an OpenAI API key
• Supports chat completion format

Claude

• Uses Claude-2 model
• Requires an Anthropic API key
• Supports completion format

Error Handling

The library uses anyhow for error handling. All errors are wrapped in anyhow::Result, making it easy to handle different types of errors that might occur during API calls.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

This project is licensed under the MIT License - see the LICENSE file for details.