Open-source Langflow alternatives

How the open-source Langflow alternatives differ

Langflow is a visual builder for AI agents and workflows: you connect components on a canvas, and each flow ships as an API or an MCP server. A Langflow alternative is any tool that does that job instead, and the four rivals compared on this page sort into three shapes.

LangGraph is a code framework you assemble in Python or TypeScript. Flowise and Dify are self-hostable platforms you click through. n8n is a workflow tool with AI steps. Kortix is a fourth shape: a place where the agents run on their own cloud computers.

Kortix

Kortix puts your agents, their skills, your company memory, and every connector in one git repo you own, and runs the agents on real cloud computers. It suits teams that want agents doing finished work, like a report, a fix or a reply, with the configuration in files they can read and a person approving each change.

Kortix has 20,000+ GitHub stars (checked October 2026). Its source is on Kortix on GitHub. Open source. Read it, fork it, audit it.

One isolated sandbox per session. Each session has its own isolated machine and branch. Session work reaches main through a change request, and merge is default-deny for agents. Agents reach 3,000+ apps in a click, plus MCP, OpenAPI, Postman, GraphQL and raw HTTP, and connector credentials are brokered server-side and never enter the machine.

The fastest path to a first session is three commands:

1curl -fsSL https://kortix.com/install | bash2kortix init3kortix ship

Run it on Kortix Cloud, in your VPC, or on your own on-prem network. Self-host is free.

LangGraph

LangGraph is a low-level framework from LangChain Inc for building long-running, stateful agents, and it is a library rather than a hosted product. It suits engineers who want to write the orchestration themselves and keep durable execution, human-in-the-loop interrupts and memory inside their own code. The repository carries about 43,300 stars and shipped a release in October 2026, and its source terms are permissive (langchain-ai/langgraph, checked October 2026). LangGraph itself is free; the paid part is LangSmith Deployment, the platform LangChain sells for deploying agents.

Kortix sets out the head-to-head in Kortix vs LangGraph.

Flowise

Flowise is a drag-and-drop builder for LLM apps and multi-agent systems, and its repository is archived: active development stopped in July 2026 and the GitHub project moved to a public archive in August 2026. It suits a team that already runs a Flowise flow and can keep a fork alive, on the understanding that no upstream releases are coming. The repository carries about 55,500 stars, its last release was in July 2026, and the project stays readable and forkable (FlowiseAI/Flowise, checked October 2026). Flowise also sold Flowise Cloud and an enterprise tier.

Kortix compares the two in Kortix vs Flowise.

Dify

Dify is an open-source platform for LLM apps: a visual workflow canvas, RAG pipelines and agents in one workspace. It suits teams that want to build and run many AI apps from one platform, with a cloud edition, a self-hosted community edition and an enterprise edition. Dify carries about 158,600 stars and shipped a release in September 2026, and its own terms add two conditions to an open-source base: no multi-tenant hosting without permission, and the brand staying in place (langgenius/dify, checked October 2026). Dify Cloud starts on a free tier, and self-hosting runs in Docker Compose.

Kortix sets out the head-to-head in Kortix vs Dify.

n8n

n8n is a workflow tool with AI steps, and of the four rivals it is the closest to what Langflow's canvas does. It suits teams that want to replace point-to-point automation, with 1,500+ integrations and a visual canvas they can extend in code. n8n calls its model fair-code rather than open source: the source is public and self-hostable, and its terms restrict hosting n8n as a service for others (n8n-io/n8n, checked October 2026). The repository carries about 207,500 stars and shipped a release in October 2026.

Which open-source Langflow alternative fits

Pick LangGraph when you want a framework and will write the orchestration yourself. Pick Flowise only when an archived builder is acceptable and you can carry a fork. Pick Dify when one platform for workflows, RAG and agents is the goal.

Pick n8n when the job is a fixed workflow with integrations. Pick Kortix when the job is open-ended work an agent should finish on its own cloud computer, with the configuration in a git repo you own and each change landing for a person to review. The comparison table on this page sets the five side by side.