zavora-ai/adk-rust

Rust Agent Development Kit (ADK-Rust): Build AI agents in Rust with modular components for models, tools, memory, realtime voice, and more. ADK-Rust is a flexible framework for developing AI agents with simplicity and power. Model-agnostic, deployment-agnostic, optimized for frontier AI models. Includes support for real-time voice agents.

What it solves

ADK-Rust provides a production-ready, type-safe framework for building and orchestrating AI agents in Rust. It addresses the complexity of moving from simple LLM prompts to autonomous agents that can maintain state, execute complex workflows, handle errors, and integrate with a wide variety of models and external tools without being locked into a single provider.

How it works

The framework is model-agnostic and async, utilizing a modular system of 43 crates. It allows developers to define agents using a variety of templates (such as sequential, parallel, or graph-based workflows) and extend them with "addons" for telemetry, authentication, and memory. It supports a wide range of LLM providers including Gemini, OpenAI, Anthropic, and local inference via mistral.rs. For complex logic, it uses graph workflows with checkpoints for durable execution and human-in-the-loop approvals.

Who it’s for

It is designed for developers building enterprise-grade AI agents who require the performance and safety of Rust, specifically those needing durable execution, high-concurrency orchestration, and deployment flexibility across cloud platforms like Google Cloud Vertex AI.

Highlights

  • Model Agnostic: Supports a vast array of providers including Gemini, OpenAI, Anthropic, DeepSeek, Groq, Ollama, and Bedrock.
  • Durable Execution: Graph workflows with SQLite checkpointers allow agents to pause and resume exactly where they stopped.
  • Extensive Tooling: Includes a built-in runtime UI for visualizing conversations and agent topology, as well as a CLI tool (cargo-adk) for scaffolding projects.
  • Enterprise Ready: Features built-in support for RAG, semantic memory, OpenTelemetry tracing, RBAC, and agentic commerce (payments).
  • High Performance: Boasts sub-millisecond agent-loop overhead.

Related

  • Project
    ldclabs/anda

    A Rust framework for building composable AI agent runtimes that integrate models, tools, and memory into a single orchestrated execution environment.

  • Project
    openai/openai-agents-python

    A lightweight, provider-agnostic framework for building multi-agent workflows, supporting text, voice, and sandboxed execution across various LLMs.

  • Project
    liquidos-ai/AutoAgents

    A production-grade multi-agent framework in Rust that provides a type-safe environment for building intelligent systems with structured tool calling and pluggable LLM backends.

  • Project
    a-agmon/rs-graph-llm

    A stateful graph workflow framework for AI agents in Rust that enables the creation of complex, resumable, and type-safe agentic workflows with human-in-the-loop capabilities.

  • Project
    eastreams/loong

    A Rust-based foundation for vertical AI agents focusing on security, performance, and extensibility through an Actor runtime and action-based authorization.