Sim is the open-source AI workspace where teams build, deploy, and manage AI agents. Connect models and tools, store shared data, and inspect what happens when an agent runs.
Choose how to build
| Method | Use it to |
|---|---|
| Chat | Describe what you need and ask Sim to create or edit workspace resources. Open what it creates, run it, and check the result. |
| Workflow builder | Connect blocks into agent logic, configure their inputs, and inspect execution step by step. |
| API and SDKs | Create resources and run deployed workflows from your application. |
You can start in Chat and refine the same workflow in the builder. On self-hosted deployments, Chat depends on the deployment's configuration and can be disabled.
Follow Getting started to build an agent with web search tools and structured output.
Workspace resources
A workspace holds the resources your team and agents use:
| Resource | Purpose |
|---|---|
| Workflows | Executable logic made of connected blocks. Use Agent blocks for model reasoning and tools for actions such as sending email or querying an API. |
| Tables | Structured rows that workflows can read, update, and process. |
| Files | Documents and media you upload, create, and share with agents. |
| Knowledge bases | Searchable documents. Retrieve relevant passages for an agent to use in its answers. |
| Logs | Run records and block inputs, outputs, errors, timing, and costs. |
Deploy a workflow to make it available through an API, a chat interface, or an MCP server. MCP (Model Context Protocol) lets AI applications discover and call tools exposed by a server.
Models and integrations
Use the integration catalog to connect supported models, communication services, productivity apps, search tools, and databases. These tools let agents research a topic, process documents, update business systems, and return results to another application.
For tools outside the catalog, connect an external MCP server or create a custom tool. Configure the credentials and permissions each service requires.
Choose a deployment
- Sim Cloud: Sign in at sim.ai to use the managed service.
- Self-hosted: Run Sim on your infrastructure with Docker Compose or Kubernetes. You can connect local models through Ollama or vLLM. Chat and external integrations can still require network access.
The core source is available on GitHub under the Apache 2.0 license. Enterprise features use a separate license that requires an active Enterprise subscription for production use. See self-hosting for installation requirements and configuration.
Related docs
- Getting started: Build and test a research agent.
- Chat: Build and manage workspace resources conversationally.
- Workflows: Blocks, connections, triggers, and execution.
- Quick reference: Look up builder actions.