Runs stateful AI agents, workflows, and sandboxed code as Rivet Actors with SQLite storage on S3, for developers building durable agent backends at scale.
Use it now
Keep it with up to 9 other tools you're setting up, then share them all as one link.
Try this first
Ask your assistant, with Rivet connected:
- Durable Actors
Use Rivet to create long-lived actors: ‹what you need›
- Workflows
Use Rivet to run multi-step operations on actors: ‹what you need›
- agentOS Sandbox
Use Rivet to run the code an agent writes: ‹what you need›
Connect it over MCP
{
"mcpServers": {
"rivet-dev": {
"url": "https://mcp.rivet.dev/mcp"
}
}
}claude mcp add --transport http rivet-dev https://mcp.rivet.dev/mcp
npx -y @rivet-dev/mcp --target local
In Grok, ChatGPT and other assistants, add a custom connector with the address https://mcp.rivet.dev/mcp (how, step by step).
It may ask you to sign in to Rivet or for a key when you add it; the setup guide says which.
Setup guide: rivet.dev/docs
Call the API
Auth A bearer token from Rivet
Setup guide: rivet.dev/docs
Compare
Press Compare to add it to the tray at the bottom of the window, then pick up to 2 others anywhere on the site.
- Access
- APIMCP
- Auth
- Bearer Token
- Skills
- 5
Skills
Durable Actors
Creates long-lived actors with in-memory state, a per-actor SQLite database, WebSockets, and scheduling, one per agent, user, or session.
Workflows
Runs multi-step operations on actors that replay from where they stopped instead of starting over after a restart.
agentOS Sandbox
Runs the code an agent writes in a Linux-compatible JS and Python sandbox with control over network, filesystem, and resource limits.
Manage Actors over MCP
Lists running actors, calls actor actions, and opens the Actor Inspector from an MCP client through Rivet's hosted MCP server.
Realtime Broadcasts
Broadcasts events to every connected client over WebSockets, such as streaming model tokens to all users of one agent.
