Plans checked on vendor pages · November 11, 2026
Eight tools that let you build an AI agent without writing code. Each one is priced differently, by executions, credits, conversations or runs, and each one is good at a different job. Tell us yours and we point you to the right one.
Agentic AI vs generative AI
Generative AI writes content when you ask. Agentic AI pursues a goal over several steps: it plans, uses your apps, checks the result and keeps going until the job is done.
Free tier, paid entry plan and billing unit as published by each vendor on November 11, 2026. Open-source projects show their GitHub stars on the same date.
| Builder | Billed by | Free | Paid from | Self-host |
|---|---|---|---|---|
| n8n | Workflow executions | Community Edition, free to self-host | Cloud Starter 20 euros/mo (annual) for 2,500 executions; Pro 50 euros/mo for 10,000 | Yes |
| Zapier Agents | Activities | Free: 400 activities a month | Pro $33.33/mo billed annually ($50 monthly) for 1,500 activities; Advanced $150/mo for 5,000 | No |
| Lindy | Credits | $50 in credits for 7 days, no card | From $49.99/mo per active user for 5,000 credits; top-ups at $10 per 1,000 credits | No |
| Gumloop | Credits | 14-day free trial | Pro starts at $37/mo | No |
| Botpress | Conversations | Free: 100 conversations a month, 3 seats | Plus $150/mo billed annually for 250 conversations; Team $750/mo for 1,500; extra packs of 100 | No |
| Dify | Message credits | Open-source edition free to self-host | Cloud Professional $590 per workspace per year for 5,000 message credits a month; Team $1,590 per year for 10,000 | Yes |
| Flowise | Self-hosted | Open source, free to self-host | Hosted cloud plans on the vendor site | Yes |
| StackAI | Runs | Free: 500 runs a month, 1 member, 2 projects | Enterprise on custom quote, with SSO, on-premise and BAA options | No |
Open source · n8n-io/n8n, about 210k GitHub stars (fair-code licence)
Best for. Agents that must act across many business apps, with full control over data and hosting. AI Agent node, MCP and AI nodes on every plan.
Watch out. More technical than the others: you design the workflow around the agent yourself.
Best for. Teams already on Zapier who want agents that use its large catalogue of app connections, with no setup.
Watch out. Billed separately from Zaps, and costs climb with volume.
Best for. Personal and team assistants for email, meetings, scheduling and follow-ups, set up in plain language.
Watch out. Every active user is billed, and bigger jobs use more credits.
Best for. Visual AI workflows for marketing and operations teams: scraping, enrichment and content pipelines.
Watch out. Credit-based, so heavy AI steps use your allowance faster.
Best for. Customer-facing chat agents on your website and messaging channels, with branding and live-chat handover.
Watch out. Priced per conversation, which suits support bots more than back-office agents.
Open source · langgenius/dify, about 161k GitHub stars
Best for. Building AI apps and agents with RAG over your documents, for developers and technical teams.
Watch out. Strongest when someone on the team can manage models, knowledge bases and hosting.
Open source · FlowiseAI/Flowise, about 56k GitHub stars
Best for. A drag-and-drop builder for LLM agents and chat flows that you run on your own server.
Watch out. You maintain hosting, updates and model keys yourself.
Best for. Enterprise and regulated teams that need agents with compliance features and private deployment.
Watch out. Beyond the free plan, pricing is quote-only.
The agents that pay for themselves do one repetitive job with clear rules, inside the tools you already use, with a person approving anything sensitive.
Reads every inbound enquiry, asks the missing questions, scores it against your criteria and books the good ones.
Answers common questions from your help content and routes the rest to the right person with a summary.
Drafts replies, chases unanswered emails and keeps the CRM updated after every exchange.
Pulls the fields you need out of invoices, contracts and forms, and flags anything uncertain for review.
Collects numbers from several tools each week and writes the summary your team actually reads.
Watches competitors, reviews or mentions and sends a short digest with sources.
See each of these as a full workflow, problem and solution, in our catalogue of 175 automations.
Six steps that work in any of the builders above. Most failed agents skip step one or step five.
We build AI agents on n8n and Claude, in your own accounts, with human approval on sensitive steps and 30 days of tuning after launch. Fixed price, scope agreed in writing.
A tool that lets you create an AI agent, software that reads a request, decides what to do and acts in your apps, by configuring it visually or in plain language instead of writing code.
It depends on the job. For agents that act across many business apps with control over hosting, n8n. For quick assistants in plain language, Lindy or Zapier Agents. For website chat, Botpress. For apps over your documents, Dify or Flowise. Use the picker on this page with your own needs.
Most tools on this page have a free plan or trial. Paid plans start from about $37 to $50 a month for small teams, with usage-based limits such as credits, conversations or executions. Self-hosted open-source options cost only your server and AI model usage. Prices checked November 11, 2026.
Generative AI creates content: text, images or code in answer to a prompt. Agentic AI uses a model to pursue a goal over several steps: it plans, calls tools, checks the result and continues. Most business agents combine both: generation inside a loop of actions.
They are when their permissions are narrow and sensitive actions need human approval. Give an agent access only to the tools its job needs, log what it does, and review its output before it reaches customers on anything that matters.
Yes. Every tool on this page has a visual or plain-language builder. Coding helps for custom integrations, but most business agents can be built by connecting apps and writing clear instructions.