IBM Langflow

What it does

IBM Langflow is a visual builder for AI workflows. You assemble flows from drag-and-drop components that connect to APIs, internal services, vector databases, LLM providers and MCP tools, and you can extend any component with Python. A finished flow can be exposed as a REST API, wrapped in a custom interface, or run as an agent that calls tools and databases.

People typically start in Langflow Desktop or from a template, connect components, then test in the Playground, where they can inspect each component's response and logs. To deploy, they enable API access in the flow settings, copy a code snippet, add a Langflow API key, and run it via Docker, a remote server, Langflow Cloud or watsonx Orchestrate. The same flow definition moves between dev, QA and production. A tutorial shows a flow imported into watsonx Orchestrate as a tool.

It suits developers and technical teams prototyping agent applications, from low-code to full-code. Desktop is offered for macOS and Windows; the open-source edition is self-managed, and IBM Elite Support is sold separately for enterprises needing SLAs. Flows still need external model providers or local models and their credentials. Pricing for the managed options is not published on the page.

What it can do

• Assemble AI workflows visually with drag-and-drop components
• Extend components and flows with Python and custom logic
• Connect to APIs, vector databases, LLM providers and MCP tools
• Test flows in a Playground with component responses and logs
• Expose flows as REST APIs or run them as tool-calling agents
• Build custom MCP servers from flows

Works with

• watsonx Orchestrate
• Langflow Cloud
• Docker
• MCP tools

Example workflows

Prototype and deploy a flow as an API

Inputs: a template or blank project plus model provider credentials. Steps: add and connect components, test in the Playground, enable API access in flow settings, copy the code snippet and add a Langflow API key. Output: a flow served as a REST endpoint via Docker or a remote server.

Package a flow as a watsonx Orchestrate tool

Inputs: a Langflow flow built with search and a local Granite model. Steps: export the flow as JSON, list dependencies in requirements.txt, import it with the ADK, then attach it to an agent. Output: an agent that calls the flow on request.

Before you choose

• Several editions exist (OSS, Desktop, Orchestrate integration); features and support differ by edition.
• Flows rely on external LLM providers or local models that you configure yourself.
• Enterprise SLAs require the separate IBM Elite Support offering.

Listed under

• Coding
• App builders
• MCP and agent tools
• Agent orchestration
• Agent workflows
• AI agents
• Developers
• Enterprise
• Enterprises

Get started

Quick start docs cover OSS, Desktop and Orchestrate setup.

1. Install Langflow Desktop (macOS or Windows), the open-source version, or use the watsonx Orchestrate integration.
2. Create a project or start from a template in the Store, then add and connect components.
3. Test in the Playground, then enable API access, add your Langflow API key and deploy.

Prices for Langflow Cloud, Elite Support and the watsonx Orchestrate integration were not confirmed; the page directs readers to contact IBM.

We couldn’t find published prices for IBM Langflow. See their pricing page ↗

Sources for this overview

Product pages, documentation, and pricing checked 29 Oct 2026. Each source supports part of the overview; availability and prices can change.

• IBM Langflow ↗www.ibm.com
• IBM Langflow ↗www.ibm.com
• Build an AI agent with Langflow, Granite 4.0 models, and watsonx Orchestrate ↗developer.ibm.com

Company website

ibm.com: product, pricing and policy pages

YC directory mirror

Batch and company name

Web search

Confirms the official site

Social accounts

Accounts the company links, and its posts on X

Human review

Checked by a person

knowledge_base

Last checked 29 Oct 2026. Every field keeps its source and date; unknown means we couldn’t confirm it.