Introduction

ai-lib is a production-grade Rust crate providing a unified, reliability-focused multi‑provider AI SDK. It eliminates the complexity of integrating with multiple AI providers by offering a single, consistent interface.

Announcement: v0.3.4 released — provider failover support, major provider expansion, enhanced multimodal content creation, and new import system. For advanced enterprise capabilities, explore ai-lib-pro with advanced routing, observability, security, and cost management features.

Goals

• Reduce integration cost across 20+ AI providers
• Improve success rate & tail latency via built-in reliability primitives
• Offer consistent streaming & function calling semantics across all providers
• Remain vendor-neutral and extensible with pluggable transport and metrics

Key Features

• Unified API for chat completions across all supported providers
• Streaming support with consistent delta handling (SSE + emulated fallback)
• Function calling with normalized tool schemas and policies
• Multimodal content support (text, images, audio) where providers support it
• Reliability primitives: retry with exponential backoff, circuit breaker, rate limiting
• Model management: performance-based selection, load balancing, health monitoring
• Batch processing with configurable concurrency limits
• Observability hooks for custom metrics and monitoring integration
• Progressive configuration from environment variables to explicit builder patterns

Supported Providers

ai-lib supports 20+ AI providers including OpenAI, Groq, Anthropic, Gemini, Mistral, Cohere, Azure OpenAI, Ollama, DeepSeek, Qwen, Baidu Wenxin, Tencent Hunyuan, iFlytek Spark, Moonshot Kimi, HuggingFace, TogetherAI, xAI Grok, OpenRouter, Replicate, Perplexity, AI21, ZhipuAI, and MiniMax.

Quick Start

use ai_lib::prelude::*;

#[tokio::main]
async fn main() -> Result<(), AiLibError> {
    let client = AiClient::new(Provider::Groq)?;
    let req = ChatCompletionRequest::new(
        client.default_chat_model(),
        vec![Message {
            role: Role::User,
            content: Content::new_text("Hello, world!".to_string()),
            function_call: None,
        }]
    );
    let resp = client.chat_completion(req).await?;
    println!("Answer: {}", resp.choices[0].message.content.as_text());
    Ok(())
}

Next: Read Getting Started, then see Features & Optional Modules, and explore Advanced Examples.