Introducing ADK-Rust: A Rust Implementation of Google's Agent Development Kit

[jkmaina]

Hey everyone! 👋

I'm excited to share ADK-Rust - a production-ready implementation of Google's Agent Development Kit in Rust.

Why Rust?
After working extensively with adk-python in developing an ai agent factory at zavora.ai, I wanted to bring the same powerful agent development patterns to the Rust ecosystem, targeting use cases where:

Performance is critical - Rust's zero-cost abstractions and memory safety
Deployment size matters - Single binary with no runtime dependencies
Systems-level integration - Embedded systems, edge computing, IoT
Concurrency at scale - Rust's async/await with tokio
Features
ADK-Rust maintains API parity with the Python ADK where possible:

✅ Model-agnostic - Gemini, OpenAI, Anthropic, DeepSeek support
✅ Multiple agent types - LlmAgent, SequentialAgent, ParallelAgent, LoopAgent
✅ Tool support - Built-in tools (Google Search, Code Execution) + custom tools
✅ MCP support - Model Context Protocol integration
✅ Sessions & Memory - InMemorySessionService, DatabaseSessionService
✅ Streaming - Full streaming support for real-time responses
✅ Telemetry - OpenTelemetry integration for tracing/metrics
✅ A2A Protocol - Agent-to-Agent communication

Quick Example

use adk_rust::prelude::*;

#[tokio::main]
async fn main() -> Result<()> {
    let agent = LlmAgentBuilder::new()
        .name("my_agent")
        .model(GeminiModel::new("gemini-2.0-flash")?)
        .instruction("You are a helpful assistant.")
        .build()?;

    let response = agent.run("Hello!").await?;
    println!("{}", response);
    Ok(())
}

Links
📦 Crates.io: https://crates.io/crates/adk-rust
📚 Docs: https://docs.rs/adk-rust
🌐 Website: https://adk-rust.com/
💻 GitHub: https://github.com/zavora-ai/adk-rust
Looking for Feedback
I'd love to hear from the community:

What features from adk-python would you prioritize?
Any interest in contributing or testing?
Use cases where a Rust implementation would be valuable?
This is an independent community project, not officially affiliated with Google, but designed to be compatible with the ADK ecosystem.

Thanks for reading! 🦀

[coldsmirk]

Excellent 👍👍👍

[jkmaina]

Thank you

[coldsmirk]

Thanks! I've looked into adk-rust more closely — it's incredibly comprehensive! SequentialAgent, ParallelAgent, LoopAgent, LangGraph-style graphs, realtime voice, MCP, browser automation, and even a visual Studio.

The Rust AI ecosystem really needs this. Existing options like rig are useful but lack these higher-level agent abstractions. Looking forward to seeing adk-rust thrive! 🦀🚀

[MartinSclxmedia]

Much needed and much appreciated.

[ptrxwsmitt]

To my mind it is about time to also add official rust support 👍

[tz-torchai]

This is much needed, thanks! Can Google add official support to ADK-rust?

[richardchen874-sys]

The Introducing ADK-Rust: A Rust Implementation of Google's Agent Development Kit can quietly multiply cost when tools, retries, or validation loops get involved, so it deserves first-class telemetry.

A compact provider matrix could track streaming, tool calls, structured output, usage reporting, timeout behavior, and fallback safety. That makes routing decisions easier to explain later.

I am testing a multi-model OpenAI-compatible API layer around official Chinese models, so the routing and usage-accounting parts of this are directly relevant to me. Which provider difference causes the most friction here: tools, streaming, usage reporting, or fallback behavior?