Langflow vs Windmill vs Dify

Side by side, as alternatives to Zapier. Rows that differ are highlighted; the strongest value in a row is marked. Unknown means no source yet.

Langflowlangflow-ai/langflow

Windmillwindmill-labs/windmill

Difylanggenius/dify

Zapierpaid product

Cost and coverage

Price

$0 software; you run it

Free (100 tasks per month, two-step Zaps); Professional from $19.99/month; Team from $69/month source ↗

App-to-app workflows

Partly

Yes

Scheduled automations

Partly

Yes

AI agent workflows

Partly

Yes

Webhooks

Partly

Yes

Jobs covered

2 of 4 jobs

3 of 4 jobs

All 4

Popularity and activity (from GitHub)

Stars

155,896

18,241

159,161Top

—

Gain this week

Unknown

—

Signal

Building history

—

Commits / 4 wk

377

651

775Top

—

Latest release

v1.12.7 · 2026-10-21

v1.846.1 · 2026-10-21Top

1.17.1 · 2026-09-10

—

Commits / 12 wk

1,288

1,979

2,560

—

People committing / 12 mo

158+

66+

253+

—

Issues opened / closed, 30 days

75 / 115

45 / 26

392 / 319

—

Facts, with sources on each project page

Runs on my machine?Differs

Yes, as a desktop app, a local Python package or a Docker container you host

Self-hosted with Docker Compose or Kubernetes; Windmill cloud also offered

Self-hosted with Docker Compose on your machine or server; Dify Cloud also offered

Hosted by the vendor

Code licenceDiffers

MIT

AGPLv3 (clients and OpenFlow spec Apache-2.0); enterprise code proprietary

Dify Open Source License: Apache 2.0 with extra conditions (no multi-tenant service without permission; keep the logo and copyright in the frontend); not an OSI licence

—

Model weights licenceDiffers

No separate model weights

No model weights

No separate model weights

—

Commercial useDiffers

Allowed under MIT

Free for internal use; reselling, managed service or embedding in a product needs AGPL compliance or a commercial licence

Commercial use allowed, but a multi-tenant service needs written permission and the web frontend's logo and copyright may not be removed

—

Paid services needed?Differs

Not required with local models via Ollama; cloud model API keys are optional

None required; Enterprise Edition and Windmill cloud are optional

Not required with local models via the Ollama plugin; cloud model APIs, Dify Cloud and Dify Enterprise are optional

Subscription

HardwareDiffers

Python package: dual-core CPU and 2 GB RAM minimum; multi-core CPU and 4 GB RAM recommended

Rule of thumb 1 vCPU and 1-2 GB RAM per worker, plus Postgres

Minimum 2 CPU cores and 4 GiB RAM; on macOS the docs ask for a Docker VM with 2 vCPUs and 8 GiB

—

Operating systemsDiffers

Desktop app for Windows and macOS (13+); Python package and Docker on Linux, macOS and Windows

Docker on macOS, Windows and Linux; native Windows workers supported

Docker on Linux, macOS 10.14+ and Windows (with WSL 2)

—

Setup effortDiffers

Easy; download Langflow Desktop, or one uv pip command, or one docker run

Medium, Docker Compose (three files) with Postgres

Medium; clone the repo, copy .env and run docker compose up

—

Stars over time

Switch to Growth % to compare projects of different sizes fairly.

Langflow
Windmill
Dify

Focus a point, use arrow keys, or tap for its value.

Tracked daily since 15 Oct 2026. Dashed history: estimated from public records of when people starred the repo. Lines connect observations only; gaps contain no measured points.

Growth % starts at each project's first observed point in the shown range. Zero-star starts are unavailable.

Notes for a video or post

A plain summary with today's date and a link to the sources. Check each project page before you publish.

Langflow vs Windmill vs Dify (as alternatives to Zapier)
Figures from RepoQuarry, 22 Oct 2026. Sources: https://repoquarry.dev/compare?ids=langflow,windmill,dify&product=zapier

Zapier: Free (100 tasks per month, two-step Zaps); Professional from $19.99/month; Team from $69/month
Langflow: free software; does 2 of 4 Zapier jobs
Windmill: free software; does 3 of 4 Zapier jobs
Dify: free software; does 3 of 4 Zapier jobs

- Langflow: 155,896 GitHub stars, 1,288 commits in 12 weeks, licence MIT
- Windmill: 18,241 GitHub stars, +49 in the last 7 days, 1,979 commits in 12 weeks, licence AGPLv3 (clients and OpenFlow spec Apache-2.0); enterprise code proprietary
- Dify: 159,161 GitHub stars, 2,560 commits in 12 weeks, licence Dify Open Source License: Apache 2.0 with extra conditions (no multi-tenant service without permission; keep the logo and copyright in the frontend); not an OSI licence

Where they differ:
- Runs on my machine?: Langflow Yes, as a desktop app, a local Python package or a Docker container you host; Windmill Self-hosted with Docker Compose or Kubernetes; Windmill cloud also offered; Dify Self-hosted with Docker Compose on your machine or server; Dify Cloud also offered
- Code licence: Langflow MIT; Windmill AGPLv3 (clients and OpenFlow spec Apache-2.0); enterprise code proprietary; Dify Dify Open Source License: Apache 2.0 with extra conditions (no multi-tenant service without permission; keep the logo and copyright in the frontend); not an OSI licence
- Model weights licence: Langflow No separate model weights; Windmill No model weights; Dify No separate model weights
- Commercial use: Langflow Allowed under MIT; Windmill Free for internal use; reselling, managed service or embedding in a product needs AGPL compliance or a commercial licence; Dify Commercial use allowed, but a multi-tenant service needs written permission and the web frontend's logo and copyright may not be removed
- Paid services needed?: Langflow Not required with local models via Ollama; cloud model API keys are optional; Windmill None required; Enterprise Edition and Windmill cloud are optional; Dify Not required with local models via the Ollama plugin; cloud model APIs, Dify Cloud and Dify Enterprise are optional
- Hardware: Langflow Python package: dual-core CPU and 2 GB RAM minimum; multi-core CPU and 4 GB RAM recommended; Windmill Rule of thumb 1 vCPU and 1-2 GB RAM per worker, plus Postgres; Dify Minimum 2 CPU cores and 4 GiB RAM; on macOS the docs ask for a Docker VM with 2 vCPUs and 8 GiB
- Operating systems: Langflow Desktop app for Windows and macOS (13+); Python package and Docker on Linux, macOS and Windows; Windmill Docker on macOS, Windows and Linux; native Windows workers supported; Dify Docker on Linux, macOS 10.14+ and Windows (with WSL 2)
- Setup effort: Langflow Easy; download Langflow Desktop, or one uv pip command, or one docker run; Windmill Medium, Docker Compose (three files) with Postgres; Dify Medium; clone the repo, copy .env and run docker compose up

Stars show popularity, not quality. Check each project page before publishing.

Want share images and a ready video description? Open a project's creator kit.

Numbers: GitHub, fetched nightly. Facts: read from each project's own documentation, with the source on its page. How we check. langflow-ai/langflow ↗windmill-labs/windmill ↗langgenius/dify ↗

Langflow vs Windmill vs Dify

Stars over time

Notes for a video or post

Cookie preferences

Report a problem