sim
Visual and code-based platform to build, run, and monitor AI agent workflows with integrated databases and knowledge.
A collaborative agent workspace linking LLMs, local models, databases, and enterprise services.
Sim is a collaborative platform for designing, deploying, and tracking AI agents and automated workflows. Municipal IT teams and smart city developers can use it to build data-driven administrative assistants and civic process automation. It runs in the browser, on desktop, or self-hosted inside your existing infrastructure.
What it is
Sim lets you construct AI agents using visual flow editors, code, or conversational prompts. It connects to external integrations and supports cloud models as well as local runners like Ollama and vLLM. Built-in relational tables, shared files, and vectorized knowledge stores provide persistent memory across workflows.
Who it's for
It suits municipal technical teams, civic software developers, and agency engineers who need a controllable platform for AI agent orchestration. It does not suit non-technical administrative staff looking for turn-key civic software without initial setup requirements.
Key features
Visual flow editor canvas
Construct multi-agent logic visually using a ReactFlow-powered canvas or write inline code steps.
Integrated tables and knowledge
Store structured relational data, vectorized document stores, and team files in a single unified workspace.
Model Context Protocol integration
Expose native MCP server endpoints to connect Claude Code, Codex, and external standard agents.
Modular CLI command management
Configure background jobs, isolated code sandboxes, caching, and custom authentication using CLI setup commands.
Up and running in 3 steps
Our honest verdict
What's great
- Connects directly to local models via Ollama and vLLM for air-gapped public sector deployments.
- Exposes native Model Context Protocol endpoints for standardized external AI client connections.
- Combines database tables, vector knowledge bases, and files within one workspace runtime.
- Offers an interactive setup CLI to manage Docker Compose and Kubernetes configurations.
Where it falls short
- Requires Node.js 20+, Docker, and ongoing server maintenance for self-hosted instances.
- Relies on external sandbox infrastructure like E2B or isolated-vm for code execution safety.
- Chat API keys and managed integrations require manual setting or external service credentials.
- Database backups and cache management must be handled separately by your operations team.
Pricing & license
Sim is published as free open-source software under the Apache License 2.0. Hosting costs depend on your infrastructure and LLM API usage. A managed cloud version is available at sim.ai for teams preferring a hosted subscription.
Open remote AI-training roles
768 open remote AI-training roles, updated November 14, 2026; most pay $20–$175 an hour. Zekai lists them free; the application goes to the employer’s platform.
Alternatives
Use it, or have us set it up
FAQ
Can Sim run completely on-premises without cloud AI services?
Yes, Sim supports local model orchestration using Ollama and vLLM along with self-hosted PostgreSQL databases.
How does Sim handle untrusted code execution inside workflows?
Sim isolates custom code execution using isolated-vm or cloud sandboxes like E2B.
Is there an MCP endpoint to interact with external tools?
Yes, Sim provides a dedicated MCP server URL at `/mcp` compatible with MCP clients like Claude and Codex.
