Activepieces
MIT Community Edition whose TypeScript pieces double as MCP servers, with loops, branches, approvals, code steps and an AI SDK.
Know firstEnterprise features sit in paid folders, and its README gives two different integration counts.
Workflow automation6 projects, checked September 17, 2026
Open-source alternatives to Zapier
The open-source alternatives to Zapier are n8n, Activepieces, Windmill, Huginn, Dify and Langflow. n8n has the largest integration library, with 1,500 by its own count, and Activepieces is the MIT no-code builder whose integrations double as MCP servers. Windmill suits developers who would rather write scripts, Huginn has run private rule-based automations since 2013, and Dify and Langflow build AI workflows and agents. Two, n8n and Dify, are source available rather than open source.
- A no-code builder under MIT
- Activepieces
- The largest integration library
- n8n
- Code-first automation for developers
- Windmill
4 of 6 carry an OSI licence. 6 of 6 run on your own machines. Figures read from GitHub on September 17, 2026.
Pick by need
One project per need, with the reason from its own README and the one thing to know before you choose it.
n8n
1,500+ integrations and 9,000+ templates by its own count, with AI agents, JavaScript and Python steps, and role-based access, self-hosted or in its cloud.
Know firstSource available: its licence allows internal, personal and non-commercial use only, and enterprise files need a paid licence.
Windmill
Scripts in Python, TypeScript, Go, Bash, SQL and more, composed into flows with generated UIs, triggered by schedules, webhooks, Kafka or email.
Know firstAGPLv3 from source, but the ready-made images carry extra terms and may not be resold.
Dify
Visual AI workflows, retrieval pipelines, agents with their own sandbox, MCP tools and hundreds of models, with an API for everything built.
Know firstSource available: no multi-tenant use and no logo removal without written permission.
Langflow
MIT visual builder for agents and workflows, with Python-editable components and every flow deployable as an API or MCP server.
Know firstBuilt around AI components rather than app-to-app connectors, with no integration count given.
Huginn
MIT since 2013, with agents that watch the web, send and receive webhooks, run JavaScript and email you when a page changes.
Know firstNo AI features at all, and fewer integrations than the modern platforms.
What to take away
- Activepieces is the MIT no-code pick; n8n has the largest integration library but a restrictive licence.
- Every integration in Activepieces is also an MCP server, so the same connectors serve agents.
- Read the licence first here: two of the six are source available, for different reasons.
- Zapier still wins on thousands of maintained connectors with nothing to host.
The projects, side by side
The facts that decide most choices. Each project's page carries the full panel.
| Project | Best for | Openness | Licence | Runs as | Own machines | Stars |
|---|---|---|---|---|---|---|
| Activepieces | A no-code builder under MIT | OSI | See LICENSE | Self-hosted (see its deploy guide), Hosted cloud | Yes | 24,531 |
| n8n | The largest integration library | Source available | See LICENSE | Docker, Install script, Hosted cloud | Yes | 205,074 |
| Windmill | Code-first automation for developers | OSI | See LICENSE | Docker Compose, Kubernetes (Helm) | Yes | 17,958 |
| Dify | AI-first workflows and agents | Source available | See LICENSE | Docker Compose, Hosted cloud | Yes | 156,209 |
| Langflow | Visual AI flows as APIs or MCP tools | OSI | MIT | pip (uv), Docker, Desktop app | Yes | 154,959 |
| Huginn | Private rule-based automations | OSI | MIT | Docker, Ruby install, Heroku | Yes | 49,969 |
Licence and star figures read from GitHub on September 17, 2026. Stars are shown as a dated fact and were not used to rank this page.
Each project, in turn
An MIT automation builder its README calls an open-source replacement for Zapier, with hundreds of TypeScript pieces that double as MCP servers, AI agents, human approval steps and a no-code builder.
MIT Community Edition whose TypeScript pieces double as MCP servers, with loops, branches, approvals, code steps and an AI SDK.
Know firstEnterprise features sit in paid folders, and its README gives two different integration counts.
- Licence
- See the LICENSE file
- Runs as
- Self-hosted (see its deploy guide), Hosted cloud
- Platforms
- Self-hosted, including network-gapped
- Hosted option
- Sold by the maintainers
- Last release
- 0.91.0, Sep 14, 2026
A fair-code workflow automation platform with 1,500+ integrations, 9,000+ templates, AI agents over OpenAI, Anthropic, Google or open models, and code steps in JavaScript and Python, under the Sustainable Use License.
1,500+ integrations and 9,000+ templates by its own count, with AI agents, JavaScript and Python steps, and role-based access, self-hosted or in its cloud.
Know firstSource available: its licence allows internal, personal and non-commercial use only, and enterprise files need a paid licence.
- Licence
- See the LICENSE file
- Runs as
- Docker, Install script, Hosted cloud
- Platforms
- Docker
- Hosted option
- Sold by the maintainers
- Last release
- n8n@2.39.7, Sep 17, 2026
A developer platform for scripts, workflows, background jobs and internal UIs in Python, TypeScript, Go, Bash, SQL and more, AGPLv3 when built from source, with triggers from schedules, webhooks, Kafka and email.
Scripts in Python, TypeScript, Go, Bash, SQL and more, composed into flows with generated UIs, triggered by schedules, webhooks, Kafka or email.
Know firstAGPLv3 from source, but the ready-made images carry extra terms and may not be resold.
- Licence
- See the LICENSE file
- Runs as
- Docker Compose, Kubernetes (Helm)
- Platforms
- AWS, GCP, Azure, Fly.io, Render, Hetzner, DigitalOcean and others
- Hosted option
- Sold by the maintainers
- Last release
- v1.814.0, Sep 17, 2026
An LLM app development platform with visual AI workflows, RAG pipelines, sandboxed agents, MCP tools and hundreds of models, under Apache-2.0 plus conditions that forbid multi-tenant use and removing the logo without permission.
Visual AI workflows, retrieval pipelines, agents with their own sandbox, MCP tools and hundreds of models, with an API for everything built.
Know firstSource available: no multi-tenant use and no logo removal without written permission.
- Licence
- See the LICENSE file
- Runs as
- Docker Compose, Hosted cloud
- Platforms
- Docker
- Hosted option
- Sold by the maintainers
- Last release
- 1.17.1, Sep 10, 2026
An MIT visual platform for building and deploying AI agents and workflows that turns every flow into an API or an MCP server, with Python-editable components and multi-agent orchestration.
MIT visual builder for agents and workflows, with Python-editable components and every flow deployable as an API or MCP server.
Know firstBuilt around AI components rather than app-to-app connectors, with no integration count given.
- Licence
- MIT
- Runs as
- pip (uv), Docker, Desktop app
- Platforms
- Anywhere Python or Docker runs, Desktop app for Windows and macOS
- Hosted option
- None from the maintainers
- Last release
- v1.12.2, Sep 16, 2026
An MIT system, running since 2013, for building agents that watch the web and act for you, which its README calls a hackable version of IFTTT or Zapier on your own server, with webhooks, scraping and JavaScript steps.
MIT since 2013, with agents that watch the web, send and receive webhooks, run JavaScript and email you when a page changes.
Know firstNo AI features at all, and fewer integrations than the modern platforms.
- Licence
- MIT
- Runs as
- Docker, Ruby install, Heroku
- Platforms
- Docker, Any server with Ruby
- Hosted option
- None from the maintainers
- Last release
- v2026.09.09, Sep 8, 2026
How to choose
The questions that settle it, in the order they usually come up.
Who builds the automations?
Non-technical staff: Activepieces or n8n. Developers: Windmill. AI engineers: Dify or Langflow.
Does the licence have to be OSI-approved?
Activepieces, Langflow and Huginn are MIT; Windmill's source build is AGPLv3. n8n and Dify are source available.
Do you need AI agents?
n8n, Activepieces, Dify and Langflow build them; Huginn has none.
Where Zapier is still the better choice
Zapier, in the vendor's own words, is a platform to "build and scale AI workflows and agents across 9,000+ apps", by its own description; the app count is Zapier's own claim. zapier.com. The cases below are where that still wins.
- You want thousands of app connections maintained for you, with nothing to host.
- You want non-technical staff building automations with support behind them.
- You want the vendor to handle reliability, retries and scaling.
- You want AI agents connected to your business apps with no infrastructure work.
How this page was checked
Numbers from GitHub, dated
Stars, licence, last commit and last release are read from the GitHub API by a script every week and carry the date they were read. Nobody types them.
Claims from the README
What each project does is written from its own README and LICENSE, read in full on the date shown, and phrased as the project's claim. The sources are listed at the end.
What is never said
That a project is faster, better or cheaper than Zapier. A price for anyone. A user count. A roadmap. If a fact is not on this page, it was not in the source.
- OSI Open source (OSI licence)
- The code, and the weights where the project is a model, carry a licence on the OSI approved list.
- Source available Source available
- The code is public, and the licence limits what you may do with it. Read the LICENSE file before commercial use.
- Open weights Open weights
- The model weights can be downloaded and run. The training data or code is closed, or the weights carry a restriction the code does not.
Questions people ask
What is the best open-source alternative to Zapier?
Activepieces (MIT) for a no-code builder whose pieces double as MCP servers. n8n for the biggest integration library, though its licence limits you to internal and non-commercial use. Windmill for code-first automation, Huginn for private rule-based agents, and Dify or Langflow when the workflow is mostly AI.
Is n8n free to use?
For your own internal business purposes, and for personal and non-commercial use, yes, under its Sustainable Use License. You may not offer it to others as a service or sell a product built on it, and features in files marked .ee need a paid enterprise licence in production.
Which of these can non-technical staff use?
Activepieces is built for it: its README says developers set up the tools and anyone in the organisation uses the no-code builder, with an Ask AI feature in the code step. n8n and Dify have visual canvases too. Windmill and Huginn expect code or configuration.
Do they run AI agents?
n8n, Activepieces, Dify and Langflow do. Dify gives each agent its own sandbox to run commands and handle files, Langflow deploys flows as MCP servers, and Activepieces exposes every integration as an MCP server. Huginn has no AI features at all.
Can I self-host them?
All six, yes. Most run with Docker or Docker Compose; Windmill and Dify also document Kubernetes. Self-hosting is the reason to choose them: the data and the credentials stay on your own machines.
Sources
- Zapier homepage, the vendor's own description, read September 17, 2026
- n8n README and licences, read September 17, 2026
- Activepieces README, read September 17, 2026
- Windmill README, read September 17, 2026
- Huginn README, read September 17, 2026
- Dify README, read September 17, 2026
Running one of these inside your own environment, with your own data and your own security rules, is the kind of work Reveneau does. Read how a forward deployed engagement works.