dify vs Flowise: an active platform versus an archived visual builder
This is not an even contest. Dify is an actively developed LLM application platform whose last push was September 11, 2026, while Flowise has been archived: its README announces the fact at the top and points to a Future of Flowise discussion. Flowise still works for quick visual prototypes, but anyone choosing between these two for a new system should treat only Dify as a live option.
At a glance
| Project | langgenius/dify | FlowiseAI/Flowise |
|---|---|---|
| Licence | Custom licenceCustom licence: read the LICENSE file | Custom licenceCustom licence: read the LICENSE file |
| Maintenance | Commits in the last dayLast push September 29, 2026 | ArchivedLast push August 13, 2026 |
| Language | TypeScript | TypeScript |
| GitHub stars | 157,512 | 55,492 |
| Read more | Our analysisGitHub | Our analysisGitHub |
Which one to choose
Choose dify if you need a maintained platform for visual AI workflow building, RAG pipelines, agent tools and model management, with a Docker Compose deployment, observability integrations and APIs for your own business logic.
Choose Flowise if you want a fast self-hosted visual prototype of an agent or chatflow, you are comfortable accepting that development has stopped, and you will not build a production system on it.
An archived project and an active one
The maintenance facts decide this comparison before the features do. Flowise is archived. Its README opens with an unmissable line stating that the project has been archived and directing readers to a Future of Flowise discussion for what happens next, and its last push was August 13, 2026. Dify is not archived, and its last push was September 11, 2026, with releases including 1.17.0 in August 2026. Both projects are TypeScript codebases with licences GitHub cannot classify, so neither offers the reassuring MIT or Apache note you get from many open-source tools. But the practical difference is stark: Dify keeps shipping, and Flowise does not. For a team that wants a long-lived platform, that alone settles the question, and the rest of this page explains what each still offers and where each falls short.
What each platform offers
Dify describes itself as an open-source LLM application development platform whose interface combines AI workflow, RAG pipeline, agent capabilities, model management and observability. Its feature list is broad: a visual workflow canvas, a prompt IDE for crafting prompts and comparing model performance, a RAG pipeline with out-of-box text extraction from PDFs, PPTs and other common formats, agents built on LLM function calling or ReAct with 50-plus built-in tools such as Google Search, DALL E, Stable Diffusion and WolframAlpha, LLMOps for monitoring logs and performance, and Backend-as-a-Service APIs on every offering. It also advertises integration with hundreds of models from dozens of providers, including OpenAI API-compatible models. Flowise's pitch is narrower: build AI agents visually. Its README describes a visual builder for agents and chatflows, a monorepo with a Node server, a React UI, third-party node components and auto-generated API documentation, and a cloud offering. Both are visual builders; Dify is a platform around the builder, Flowise is mostly the builder itself.
Getting each one running
The install paths show different shapes. Dify starts through Docker Compose: copy the .env.example file into the docker folder and run docker compose up -d, then open the dashboard at localhost/install and run the initialization. Its README states the minimum system requirements as 2 CPU cores and 4 GiB RAM, and notes that Docker Compose 2.24.0 or later is expected. Flowise installs as an npm package on Node, version 20 or later: npm install -g flowise, then npx flowise start, then open localhost:3000. Docker Compose and a Docker image are also documented, plus elaborate self-hosting guides covering AWS, Azure, Digital Ocean, GCP, Railway, Render and others. Building Flowise from source means cloning the monorepo, installing with pnpm and building, and the README documents a known failure mode with a JavaScript heap out of memory error and workaround, which is a preview of the friction you can expect when operating an unmaintained codebase at current toolchain versions.
Operations: what keeps running after setup
Dify's operational story is designed for a running product. The README points to monitoring and analysis of application logs and performance over time, integrations with observability tools including Opik, Langfuse and Arize Phoenix, and APIs on every offering so business logic can call the workflows. Its LLMOps angle, improving prompts, datasets and models from production data and annotations, assumes a system that keeps operating. Flowise's operational story is frozen at its archive date. The README documents Flowise Cloud and infrastructure guides, but there are no new releases to pick up security fixes or compatibility changes, and the project's own analysis says to verify whether the archived codebase still installs cleanly on your Node version. A prototype that runs today may not install at all after the next Node release. That asymmetry, not any single feature, is the operational difference that matters most.
Where each falls short
Dify's weaknesses are about weight and control. The recommended path is a multi-service Docker Compose deployment, which requires maintenance, and the adoption analysis of the project says teams that need deep customization of the underlying orchestration logic will hit the platform's limits, since you work inside its canvas rather than rewriting it. The licence is another open item: the repository does not clearly state a licence, and GitHub classifies it as a custom licence, so the analysis advises checking the licence terms against your use case before adopting. Flowise's weaknesses are simpler and more final. It is archived, so there is no maintenance, no security updates and no promise that the codebase installs on current Node versions. The analysis of Flowise says it can still make sense for a quick self-hosted visual prototype and nothing more, and recommends checking the Future of Flowise discussion for an officially recommended fork or successor before relying on the project at all.
Licence, maintenance and the choice
Both repositories carry licences GitHub cannot classify, so check the LICENSE file for each project against your use case; neither offers a familiar permissive licence out of the box. On maintenance the two cannot be compared fairly, because Flowise is archived with its last push on August 13, 2026, and its Future of Flowise discussion is the place to look for a successor, while Dify is active with a push on September 11, 2026. For concrete situations: choose Dify when you intend to run an LLM application platform for a while, can maintain a Docker Compose stack, and want visual workflows, RAG, agents, model management and APIs in one place, and verify the licence terms and your model providers before building. Choose Flowise only for a throwaway prototype, or if you are already running it and want to keep the archived version working, and verify it still installs on your Node version before depending on it for anything longer than an experiment.
Bottom line
Choose Dify for any new work, because it is the maintained option with visual workflows, RAG, agents and Backend-as-a-Service APIs, and choose Flowise only for a disposable prototype where you accept that development has stopped. Verify before you decide: for Dify, that the licence terms suit your use case and your model providers are in the supported list; for Flowise, check the Future of Flowise discussion and confirm the archived codebase still installs on your Node version.