Sim
by Sim (Sim Studio)
Open-source, Apache-2.0 AI workspace for building, deploying and governing AI agent workflows
Sim is an open-source AI workspace where teams build AI agents on a visual canvas, in chat or in code, then deploy them as chat, API or MCP endpoints. It suits engineering and operations teams that want an Apache-2.0 alternative to n8n or LangFlow they can self-host, with spend limits, SSO and audit trails for enterprise rollout.
Sim, built by Sim Studio (Y Combinator X25, founded by Emir Karabeg and Waleed Latif), is an Apache-2.0 platform for building and running AI agent workflows. Users assemble agents on a Figma-style canvas, describe them in natural language to a built-in copilot, or write them with the TypeScript and Python SDKs and a CLI. Workflows combine models from OpenAI, Anthropic, Google, xAI, Mistral and Perplexity with custom functions, knowledge bases, tables, files, schedules and event triggers, and the GitHub README lists 270+ integrations covering 5,400+ actions and 520+ triggers, including Slack, Notion, HubSpot and Salesforce. Finished agents deploy as chat interfaces, API endpoints or MCP servers, and Sim also exposes its own MCP server so Claude Code or Codex can manage workspaces. Sim Logs traces each run block by block with cost and failure detail. The company announced a $7M Series A led by Standard Capital on November 12, 2025, reporting 60,000+ developers, 100+ companies in production and 10M+ workflows processed, and naming Rivian (Volkswagen Group Technologies), eXp Realty and thinkproject as users. In March 2026 it repositioned as "The AI Workspace" with Mothership, a natural-language control plane over workflows, tables and files. The repository had about 30.1k GitHub stars when checked on November 10, 2026. Sim can be self-hosted free with Docker Compose or Kubernetes, with Ollama and vLLM for local models; SSO, access control, SOC 2 and self-hosted enterprise features require the Enterprise plan. It competes with n8n, Dify, Langflow and Flowise.
Platform or AI engineering leads who need a self-hostable, permissively licensed agent builder that business teams can also use, with central spend and access controls.
Teams prototype an agent visually and ship it as an API, chat or MCP endpoint from the same workspace, with per-member spend limits and run-level traces.
At a Glance
- Category
- Agent Development
- Pricing
- Freemium, Subscription, Usage-based
- Target Market
- CTOs, Enterprise Developers, AI Engineering Teams, Operations Teams
- Deployment
- Open-source, Self-hosted, Cloud-first
- Founded
- 2025
- Headquarters
- San Francisco, USA
- Customers
- 100+ companies in production and 60,000+ developers (company-reported, November 2025)
Key Features
- ✓Visual, chat and code builders
Build the same agent on a drag-and-drop canvas, by describing it to the copilot, or with the TypeScript and Python SDKs and CLI.
- ✓270+ integrations
Prebuilt blocks cover 5,400+ actions and 520+ triggers across tools such as Slack, Notion, HubSpot, Salesforce and databases.
- ✓Deploy as chat, API or MCP
Publish a workflow as a chat interface, a REST endpoint or an MCP server, protected by SSO, email or password access.
- ✓Sim Logs tracing
Traces every run block by block with actions, cost and failures, so teams can debug agents and attribute model spend.
- ✓Governance controls
Permission groups limit which models, integrations and tools members use, with org-wide and per-member spend limits and audit records.
- ✓Self-hosting with local models
Docker Compose or Kubernetes deployment, an npx setup wizard, and Ollama or vLLM profiles keep data and inference on your own hardware.
- ✓Knowledge bases, tables and files
Built-in storage for documents and structured rows gives agents RAG context and working data without a separate database.
Capabilities
Use Cases
- •Internal support and IT agents
IT teams connect a knowledge base and ticketing tools, then expose the agent in Slack so employees get answers and routine requests handled.
- •Sales operations automation
RevOps teams chain CRM triggers, enrichment and LLM steps so new leads are researched, scored and written back to Salesforce or HubSpot automatically.
- •Replacing LangGraph-backed APIs
Engineering teams rebuild an agent deployed as an API in Sim's canvas, gaining versioned deployments and per-run traces without maintaining orchestration code.
- •Private agents on local models
Regulated teams self-host Sim with Ollama or vLLM so prompts and documents stay inside their network while business users still build workflows.
Ideal For
Best For
- ✓Teams replacing n8n that need an OSI-approved Apache-2.0 licence for self-hosting
- ✓Deploying internal agents as API or MCP endpoints behind SSO
- ✓Running agent workflows on local models through Ollama or vLLM in a private environment
- ✓Ops, finance, HR and IT teams automating multi-step work across SaaS tools like Salesforce, HubSpot and Slack
Not Ideal For
- ✗Purely deterministic integrations with no AI steps, where commenters on Hacker News noted a classic workflow tool is simpler
- ✗Small self-hosting setups with little memory: the project advises at least 12GB RAM for the Docker stack
- ✗Buyers who need a long vendor track record or HIPAA commitments, since the company is about 18 months old and its pricing page does not mention HIPAA
Integrations
Deployment
Market & Ratings
100+ companies in production and 60,000+ developers (company-reported, November 2025)
Market Analysis
Pros
- ✓Permissive Apache-2.0 licence with free, unlimited self-hosting
- ✓Large and active open-source community, with about 30.1k GitHub stars
- ✓Polished visual canvas that HN commenters singled out, plus chat and code paths for the same workflow
- ✓Transparent published per-seat pricing for the cloud tiers
Cons
- ✗Heavy self-hosted footprint: the project advises at least 12GB RAM, which an HN user questioned
- ✗HN commenters raised lock-in and future relicensing worries about the deployment features; the founders say current code stays Apache 2.0
- ✗Nested loops are not supported in the canvas, per the founder on Hacker News
- ✗Young company and smaller integration ecosystem than n8n; the self-hosted copilot depends on a sim.ai API key per independent analysis
Pricing
Free
$0
- ✓1,000 one-time credits
- ✓1 workspace
- ✓2 concurrent executions
- ✓30-day log retention
Pro
$25 per user/month
- ✓6,000 credits/month plus 2,000/week refresh
- ✓3 workspaces
- ✓50 concurrent executions
- ✓Chat with Sim
Max
$100 per user/month
- ✓25,000 credits/month plus 4,000/week refresh
- ✓Unlimited workspaces
- ✓Sandboxes, KB live sync, Slack Connect
Enterprise
Contact for pricing
- ✓SSO and access control
- ✓SOC 2 compliance
- ✓Self-hosted enterprise features
- ✓Dedicated support
Sim Cloud is priced per user per month plus credits, where one credit equals $0.005 of usage; annual billing is 15% off. The open-source core is free to self-host under Apache 2.0, but SSO, access control, SOC 2 and self-hosted enterprise features need an Enterprise licence with unpublished pricing. Prices read from sim.ai/pricing on November 10, 2026.
Security & Compliance
Connect
Sources
This page was written from 8 sources, 5 on domains other than sim.ai.
Stay Ahead of the Curve
Weekly enterprise AI insights for technology leaders. No spam, no vendor pitches—unsubscribe anytime.
SubscribeRelated Products
Zep
Context layer for AI agents: temporal knowledge graphs that give agents governed memory of users, data and documents
Restate
Durable execution runtime that keeps AI agents and backend workflows running through crashes and failures
Mastra
Open-source TypeScript framework and platform for building, deploying and observing production AI agents and workflows
Agentrys
Agentic design automation that lets chip teams build and own a self-improving engineering workforce