# n8n — Workflow automation platform for building and running AI agents on a visual canvas, with code where needed. > Source: The Agents Index — https://theagentsindex.com/n8n (structured, researched, re-verified) > Facts last verified: 2026-09-02 n8n is a workflow automation platform for building AI agents and business automations on a visual canvas, with JavaScript or Python steps where the built-in nodes fall short. It ships 500+ app integrations and an execution log that traces every agent decision. | Fact | Value | | --- | --- | | Website | https://n8n.io | | Price | from €24/month billed monthly, €20/month billed annually | | Free tier | Yes | Self-hosted Community edition is free indefinitely and includes almost the complete feature set, without projects, SSO, environments, external secrets, log streaming, sharing or Git version control. n8n Cloud has no free plan, only a 14-day trial with Pro-level features and no credit card: 1,000 executions, 5 concurrent executions, a 180-second timeout, 1 team project, 800 AI Assistant credits plus 100 AI credits and 20 AI Builder credits; the n8n public API is excluded. ## What it does **Writes code** · Partly — Authors n8n workflows from a plain-language description; the built-in assistant only gives guidance for the JavaScript or Python you write in a Code node, and it never touches application source. > AI Workflow Builder enables you to create, refine, and debug workflows using natural language descriptions of your goals. > — https://docs.n8n.io/build/ways-of-building-workflows/ai-workflow-builder · Use AI Workflow Builder, opening paragraph **Debugs and fixes** · Partly — It diagnoses and repairs the workflows it builds and troubleshoots node execution errors; the failure it fixes is an execution, not a bug in your repository. > AI Assistant plans it, builds it in your project, runs it, and fixes what breaks until it works. > — https://n8n.io/ai/ · Advanced AI section, AI Workflow Builder card **Writes tests** · No — Native evaluations exist, but you supply the dataset and the expected outputs; no documented feature generates test cases for you. > The foundation of evaluation is running a test dataset through your workflow. This dataset contains multiple test cases. > — https://docs.n8n.io/build/integrate-ai/test-and-improve-ai-workflows/understand-why-to-test · Understand why to test, What are evaluations? **Reviews code** · No — The review feature is a human approving a pinned workflow version against a visual diff (Enterprise, Preview); nothing in the product reviews a diff for you. > Workflow reviews let your team approve a specific workflow version before it's published. > — https://docs.n8n.io/build/manage-workflows/workflow-reviews · Review workflows, paragraph under the feature-availability note **Ships it** · Partly — What it ships is the automation itself: publishing puts a workflow or agent version into production and schedules or channels then run it. It does not deploy application code. > The published version is what runs when someone chats with the agent, a channel triggers it, or a schedule fires. > — https://docs.n8n.io/build/build-and-manage-agents · Build and manage agents, Draft and published versions **Done-for-you service** · No — Every plan is software you operate, on n8n Cloud or your own infrastructure; the only vendor-delivered service on the pricing page is support, dedicated with an SLA on Enterprise. > Dedicated support with SLA > — https://n8n.io/pricing/ · Enterprise plan card, 'Everything in Business plan, plus:' list ## How much it does unattended **Runs autonomously** · Yes — Agents run a reasoning loop until they produce a result and fire from schedules or channels with nobody watching; approval steps are something you add. > An agent is an autonomous assistant you build in n8n. > — https://docs.n8n.io/build/build-and-manage-agents · Build and manage agents, opening paragraph **Multi-agent** · Yes — A supervising agent delegates to specialised agents as tools, nestable across layers; published agents can also be attached as sub-agents. > The primary agent can supervise and delegate work to AI Agent Tool nodes that specialize in different tasks and knowledge. > — https://docs.n8n.io/integrations/builtin/cluster-nodes/sub-nodes/n8n-nodes-langchain.toolaiagent · AI Agent Tool, second paragraph **Agent permissions** · Yes — Approval is wired per tool: the workflow pauses and a reviewer approves or denies through Chat, Slack, Discord, Telegram, Gmail or Outlook. Tools with no review step execute unattended. > You can require human approval before an AI Agent executes a specific tool. > — https://docs.n8n.io/build/integrate-ai/ai-examples/human-in-the-loop-for-tools · Human-in-the-loop for tools, opening paragraph ## Which models it runs on **Claude** · Yes — Dedicated Anthropic Chat Model node for agents and chains, usable with your own Anthropic key or, on Cloud, with n8n Gateway credits. > Use the Anthropic Chat Model node to use Anthropic's Claude family of chat models with conversational agents > — https://docs.n8n.io/integrations/builtin/cluster-nodes/sub-nodes/n8n-nodes-langchain.lmchatanthropic · Anthropic Chat Model, opening paragraph **GPT** · Yes — OpenAI Chat Model node plus a full OpenAI app node (assistant, text, image, audio, video operations); Azure OpenAI has its own node. > Use the OpenAI Chat Model node to use OpenAI's chat models with conversational agents > — https://docs.n8n.io/integrations/builtin/cluster-nodes/sub-nodes/n8n-nodes-langchain.lmchatopenai · OpenAI Chat Model, opening paragraph **Open models** · Yes — Ollama chat and embedding nodes run local open-weight models; OpenRouter, Hugging Face Inference, DeepSeek and Mistral nodes cover hosted open-weight options. > The Ollama Chat Model node allows you use local Llama 2 models with conversational agents > — https://docs.n8n.io/integrations/builtin/cluster-nodes/sub-nodes/n8n-nodes-langchain.lmchatollama · Ollama Chat Model, opening paragraph **Model choice** · Yes — You pick the provider and model per node or per agent, and swap it without rebuilding the workflow; a fallback model can be attached to an agent. > Swap cloud or offline models without rebuilding your workflows. > — https://n8n.io/ · Build AI agents you can actually follow, 'Build complex AI without getting boxed in' ## Where you use it **In your editor** · Partly — No n8n editor extension: the editor is a browser canvas. Cursor, VS Code and Windsurf reach the instance through its MCP server, so workflow building happens from the IDE via an agent. > n8n groups clients into CLI (Claude Code, Codex, Gemini CLI), Web (Claude.ai, ChatGPT), and IDE (Cursor, VS Code, Windsurf), and shows setup steps tailored to your choice. > — https://docs.n8n.io/connect/connect-to-n8n-mcp-server · Connect a client, OAuth setup steps **On the command line** · Yes — Two terminals: n8n CLI over the API from any machine (workflow, execution, credential, project, package, audit commands) and a Server CLI on the host for backups and licences. > n8n CLI is a lightweight command-line client that communicates with a running n8n instance through the n8n API. It can run from any machine with network access and authenticates using an API key. > — https://docs.n8n.io/connect/n8n-cli · n8n CLI, opening paragraph **In your pipeline** · Yes — Git source control with workflow diffs gates what reaches production, the n8n CLI page names CI/CD pipeline integration as a fit, and the GitHub Trigger node fires on pull request events. > With n8n, your IT teams can use the CI/CD best practices they know and love. So they can imagine, test, and launch automations faster than ever before. > — https://n8n.io/enterprise/ · Collaboration section intro, 'Where modern DevOps practices meet automation' **In a browser** · Yes — n8n Cloud serves the same editor canvas, execution logs and Chat Hub in the browser; the free-trial signup is a hosted workspace. > Hosted version also available > — https://n8n.io/ · Runs where you decide, feature list ## Whose machine it runs on **Self-hosted** · Yes — Docker, npm or Helm on your own boxes, air-gapped if required; the free Community edition needs no licence key, and Business or Enterprise features unlock with one. > Protect your data by deploying on-prem, air-gapped if you need it. Managed cloud when you don't. > — https://n8n.io/ · Runs where you decide **Open source** · Partly — Source-available under the fair-code Sustainable Use License: free to use, modify and redistribute for internal business purposes, but not to resell or host for others. Files named .ee. sit under a separate enterprise licence. > open source licenses can't include limitations on use, so we do not call ourselves open source > — https://docs.n8n.io/privacy-and-security/sustainable-use-license · License FAQs, 'Is n8n open source?' ## What it costs to run **How it meters** · Yes — Metered per workflow execution, where one execution is one full run of a workflow regardless of steps or data; users are unlimited and AI Assistant credits are a separate monthly allowance. > Pricing based on monthly workflow executions, regardless of complexity. > — https://n8n.io/pricing/ · Pricing header, under the page title **Free tier** · Yes — 14-day n8n Cloud trial with no card (Business-plan trial does need one), and the self-hosted Community edition is free indefinitely per the Compare editions docs. > No credit card is required for the n8n Cloud trial. A credit card is only required for the Business plan free trial. > — https://n8n.io/pricing/ · FAQ, 'What's included in the n8n free trial?' **API access** · Yes — Public REST API with scoped API keys on every paid plan and self-hosted, ticked for Starter, Pro and Enterprise in the comparison table; it is switched off during the free trial and can be disabled by an operator. > Interact with n8n directly using HTTP requests. > — https://docs.n8n.io/connect/n8n-api · n8n API, 'Choose your interaction method' > REST API **MCP server** · Yes — Both directions: an instance-level MCP server exposes chosen workflows to external clients, the MCP Server Trigger node turns one workflow into its own server, and the MCP Client Tool calls outside servers. > Make your n8n workflows callable by any AI platform. Enable MCP on your instance to give Lovable, Claude, and other clients direct access to the automations you choose. > — https://n8n.io/ai/ · Advanced AI section, MCP card **Bring your own key** · Yes — Your own provider credentials are the default and can be mixed per node with n8n Gateway credits (Cloud Starter and Pro only); on self-hosted the AI Assistant is BYOK, so no credit ceiling applies. > To use your own account instead, select Use my own credential and pick or create a credential. > — https://docs.n8n.io/build/understand-workflows/use-gateway-credits · Use Gateway credits on a node ## Buying it for a team **A company can buy it** · Yes — Business and Enterprise are company purchases with invoice billing and sales-assisted procurement; Starter and Pro are self-serve. > For companies with < 100 employees needing collaboration and scale. > — https://n8n.io/pricing/ · Business plan card, description under the plan name **Seat model** · No — Not sold per seat. Users are unlimited on every plan and the quota is executions, so there is no seat floor or ceiling to negotiate; shared projects are what step up by tier (1, 3, 6, unlimited). > All plans include unlimited users & workflows and every integration. > — https://n8n.io/pricing/ · Pricing header, under the page title **Pooled budget** · Yes — Metering is per instance, not per person: one prepaid Gateway credit balance, one monthly execution quota and one AI-credit allowance shared by every user on the instance. > n8n routes the requests through its own gateway and bills the usage from your instance's prepaid credit balance. > — https://docs.n8n.io/build/understand-workflows/use-gateway-credits · Use Gateway credits, opening paragraph **Admin controls** · Yes — RBAC at instance and project level (project roles on all Cloud plans, custom roles on Enterprise), plus instance policies to enforce 2FA and block specific nodes. > Protect workflows and data with granular, custom project roles that set exactly what each user, team, or API can view, edit, publish, deploy, or manage, separating editing from publishing. > — https://n8n.io/enterprise/ · Governance & Security, 'Define fine-grained access rules per project' **Audit log** · Partly — Audit events and log streaming to a SIEM are Enterprise: the pricing comparison table shows a cross against audit logging and log streaming for Starter and Pro. > Audit events cover activations, variable changes, and user management, and Syslog streaming supports TLS. > — https://n8n.io/enterprise/ · Governance & Security, 'Stream logs and traces to your observability stack' **Single sign-on** · Partly — SAML and LDAP start on the Business plan and continue on Enterprise; the comparison table crosses out SSO for Starter and Pro. Self-hosted docs cover SAML and OIDC configuration. > SSO, SAML and LDAP > — https://n8n.io/pricing/ · Business plan card, 'Everything in Pro plan, plus:' list ## What happens to your code **Opt out of training** · Yes — n8n states the context sent to its AI features is never used to train provider models and is deleted after 30 days; admins can also stop sending workflow data at all, and telemetry is opt-out. > The data will not be incorporated, used, or retained to train the models of the AI services. Any data will be deleted after 30 days. > — https://docs.n8n.io/privacy-and-security/privacy · AI in n8n, 'How n8n uses AI' **Data residency** · Yes — n8n Cloud holds data in the EU, in Frankfurt, with no alternative region published; self-hosting puts it in whichever jurisdiction you deploy to. > For hosted plans, data is stored within the EU, on servers located in Frankfurt, Germany. > — https://n8n.io/pricing/ · FAQ, 'Is n8n secure? Where is my data stored?' **Getting out** · Partly — Workflows leave as JSON from the editor, the Server CLI (workflows, credentials, whole entities) or .n8np packages; separately, the pricing FAQ says the monthly AI credit balance refreshes and unused credits do not roll over. > n8n saves workflows in JSON format. You can export your workflows as JSON files or import JSON files into your n8n library. > — https://docs.n8n.io/build/manage-workflows/export-and-import · Export and import, opening paragraph **Certifications** · Partly — SOC 2 audited with external pen tests and a public Trust Center; GDPR is covered by a pre-signed DPA carrying the Standard Contractual Clauses. No ISO 27001 claim appears anywhere on the site. > We are SOC 2 audited and regularly perform external pen tests. > — https://n8n.io/enterprise/ · Governance & Security, 'Operate on a SOC 2 audited platform' ## Platform record (as of 2026-08-30) Reported by the code-hosting platform, not by the vendor. | Fact | Value | | --- | --- | | Stars | 202838 | | Forks | 60463 | | Last push | 2026-08-30 | ## Provenance 1 of 23 sources behind the facts above are independent of the vendor. ## Coverage elsewhere Pages about this product that the vendor does not control. The label states how independent each one is. | Source | Kind | Read | | --- | --- | --- | | en.wikipedia.org | Platform record | 2026-08-23 | | trendingtopics.eu | Independent source | 2026-08-23 | ## What changed - 2026-09-02 — Capabilities · Commercial terms · Free tier - 2026-08-31 — Capabilities · Commercial terms · Free tier · Price - 2026-08-23 — Alternatives · One-liner · Summary ## Correcting this record Anything marked "Not established" is a question we have not answered — it is not a No. If you represent n8n, or anything here is wrong or out of date, you can supply it with a source: the operations this site accepts from an agent are listed at https://theagentsindex.com/.well-known/agents.json — that catalogue needs no key to read, and each entry states its own auth. Supplying evidence is free. Nothing that can be bought here changes what a field says, whether a gap is closed, or how this record is scored.