Modern, type-safe Rust SDK for the Edgee AI Gateway.
Add this to your Cargo.toml:
[dependencies]
edgee = "2.0"
tokio = { version = "1", features = ["full"] }use edgee::Edgee;
#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {
let client = Edgee::from_env()?;
let response = client.send("gpt-4o", "What is the capital of France?").await?;
println!("{}", response.text().unwrap_or(""));
// "The capital of France is Paris."
Ok(())
}The send() method makes non-streaming chat completion requests:
let response = client.send("gpt-4o", "Hello, world!").await?;
// Access response
println!("{}", response.text().unwrap_or("")); // Text content
println!("{:?}", response.finish_reason()); // Finish reason
if let Some(tool_calls) = response.tool_calls() { // Tool calls (if any)
println!("{:?}", tool_calls);
}The stream() method enables real-time streaming responses:
use tokio_stream::StreamExt;
let mut stream = client.stream("gpt-4o", "Tell me a story").await?;
while let Some(result) = stream.next().await {
match result {
Ok(chunk) => {
if let Some(text) = chunk.text() {
print!("{}", text);
}
if let Some(reason) = chunk.finish_reason() {
println!("\nFinished: {}", reason);
}
}
Err(e) => eprintln!("Error: {}", e),
}
}- ✅ Type-safe - Leverages Rust's powerful type system
- ✅ Async/await - Built on tokio for efficient async operations
- ✅ OpenAI-compatible - Works with any model supported by Edgee
- ✅ Streaming - First-class support with
Streamtrait - ✅ Tool calling - Full support for function calling
- ✅ Zero-cost abstractions - Efficient implementation with minimal overhead
For complete documentation, examples, and API reference, visit:
👉 Official Rust SDK Documentation
The documentation includes:
- Configuration guide - Multiple ways to configure the SDK
- Send method - Complete guide to non-streaming requests
- Stream method - Streaming responses guide
- Tools - Function calling guide
Run the examples to see the SDK in action:
export EDGEE_API_KEY="your-api-key"
cargo run --example simple
cargo run --example streaming
cargo run --example toolsLicensed under the Apache License, Version 2.0. See LICENSE for details.