Rivet¶

What it is¶

Rivet is an open-source visual AI programming environment and TypeScript library developed by Ironclad. It allows developers to build, test, and debug complex multi-agent AI systems using a node-based editor. As of July 2026, it has fully integrated with the MCP 3.0 Task Protocol and Gemma 3 for high-performance visual reasoning and autonomous multi-agent coordination.

What problem it solves¶

It provides a powerful visual interface for designing AI logic, making it easier to manage complex flows and collaborate on agentic behaviors. It solves the performance and cost bottlenecks of traditional sandboxed environments through agentOS, which uses Wasm and V8 isolates for near-instant cold starts. Additionally, Rivet Actors address the need for stateful, distributed agent execution with million-scale isolated databases via SQLite for Rivet Actors.

Where it fits in the stack¶

Framework / Visual Orchestrator / Agent Runtime / Edge Infrastructure.

Typical use cases¶

Visual Agent Design: Designing intricate logic for autonomous or semi-autonomous AI agents using a node-based editor.
Stateful Edge Computing: Deploying millions of isolated, stateful actors that run at the edge with built-in SQLite persistence.
High-Performance Sandboxing: Running untrusted AI-generated code in agentOS with ~6ms cold starts, significantly faster than traditional Docker-based sandboxes.
Agentic Visual Reasoning: Leveraging Gemma 3 for processing complex visual inputs within agentic graphs.

Strengths¶

Developer-Centric Debugging: Real-time visual inspection of prompt chains and agent execution.
Extreme Performance: agentOS provides a full POSIX environment that is 32x cheaper and significantly faster than traditional VMs.
Stateful Concurrency: Native support for stateful actors using the Rust SDK or Effect SDK for Rivet Actors.
FastMCP 3.0 Integration: Built-in support for the latest Model Context Protocol for seamless tool and context sharing.

Limitations¶

Visual Overhead: For extremely simple prompt calls, the visual graph overhead may be unnecessary.
Ecosystem Velocity: The rapid shift towards a Rust-based core and Actor model requires keeping up with frequent breaking changes in the SDKs.

When to use it¶

When building sophisticated AI agents that require complex logic, state management, and durable workflows.
When you need a high-performance, low-cost sandbox for executing AI-generated code.
When you want to deploy stateful AI services at the edge that scale to zero.

When not to use it¶

For trivial, single-prompt AI tasks.
If you prefer purely code-based orchestration without any visual design or debugging components.

Getting started¶

Installation¶

To use Rivet in your project:

npm install @ironclad/rivet-node

Rivet Actors Setup¶

To create a new stateful actor using the Rust SDK:

cargo add rivet-actor

Local Development¶

Download the Rivet desktop application from the Official Website to start building graphs visually.

CLI examples¶

Running a Graph via CLI¶

rivet run my-project.rivet-project --graph "Main Graph" --input userInput="Hello AI"

Deploying to Rivet Compute¶

rivet deploy --actor my-agent-actor

Running a Rivet Actor locally¶

rivet-actor run --port 8080

API examples¶

Running a Graph in Node.js¶

import { runGraph, loadProject, NodeId } from '@ironclad/rivet-node';

async function runRivetGraph() {
  const project = await loadProject('path/to/project.rivet-project');

  const results = await runGraph(project, {
    graph: 'Main Graph' as NodeId,
    inputs: {
      userInput: { type: 'string', value: 'Hello Rivet!' }
    },
    openAiKey: process.env.OPENAI_API_KEY,
  });

  console.log(results.output.value);
}

Creating an Actor (Rust SDK)¶

use rivet_actor::prelude::*;

#[actor]
async fn my_actor(ctx: Context, input: String) -> Result<String> {
    let state: MyState = ctx.get_state().await?;
    let response = ctx.call_llm("gpt-4o", input).await?;
    Ok(response)
}

Langflow — Visual workflow builder.
Flowise — Node-based UI for LLM flows.
AG2 — Multi-agent conversation framework.
Promptfoo — Evaluation and testing for Rivet graphs.
LangGraph — Code-centric multi-agent orchestration.
PydanticAI — Type-safe agent framework from Pydantic.
Temporal — Durable execution often compared with Rivet Workflows.
Claude Code — Supported via Sandbox Agent SDK integration.

Sources / References¶

Contribution Metadata¶

Last reviewed: 2026-07-21
Confidence: high