# Langflow — Open-source visual builder for AI agents, RAG apps and MCP servers, with an API server included.

> Source: The Agents Index — https://theagentsindex.com/langflow (structured, researched, re-verified)
> Facts last verified: 2026-09-25

Langflow is an open-source, Python-based framework for building AI agents and RAG applications through a drag-and-drop visual editor. Every flow it builds is served as an API endpoint or an MCP server from a Langflow instance you run yourself.

| Fact | Value |
| --- | --- |
| Website | https://www.langflow.org |
| Price | Langflow is completely open source and you can deploy it to all major deployment clouds. |
| Free tier | Yes |

The whole application is MIT-licensed and free to install and run without limit (`uv pip install langflow`, Docker image, or desktop app). Costs you bear are your own hosting and the model provider API keys you configure; the vendor publishes no paid plan.

## What it does

**Writes code** · Yes — Langflow Assistant turns a natural-language prompt into whole flows or into Python for a single custom component, scoped to Langflow artifacts rather than an application codebase.

> Langflow Assistant understands the structure of the Langflow graph, so it can build complete flows or create individual components from natural language prompts.
> — https://docs.langflow.org/langflow-assistant · Langflow Assistant, intro, third paragraph

**Whole-codebase work** · No — The vendor bounds the assistant's context to the single flow open on the canvas, so it cannot carry a change across a repository or even across flows.

> The language model in Langflow Assistant only has the currently opened flow in your workspace for context.
> — https://docs.langflow.org/langflow-assistant · Langflow Assistant, paragraph beginning 'Behind the scenes, Langflow Assistant runs a built-in Langflow flow'

**Done-for-you service** · No — You install and operate the software yourself; Langflow offers professional services and premier support only as enquiries through a contact form, not as a managed build-and-run engagement with published scope.

> Have a question, need premier support, or want to learn more about our professional services? We're here to help.
> — https://www.langflow.org/contact-us · Page heading section, introductory sentence under "We're here for you!"

## How much it does unattended

**Runs autonomously** · Yes — Agents built in Langflow use the LLM as a reasoning engine to choose and run tools without a human choosing each step, and the docs describe this as autonomous task execution.

> Agents extend Large Language Models (LLMs) by integrating tools, which are functions that provide additional context and enable autonomous task execution.
> — https://docs.langflow.org/agents · "How do agents work?" expandable section, first paragraph

**Multi-agent** · Yes — One agent can call another agent as a tool, so several agents coordinate inside a single flow; the project advertises multi-agent orchestration as a headline feature.

> An Agent component can use any Langflow component as a tool, including other agents and MCP servers.
> — https://docs.langflow.org/agents · "Agent component parameters" > "Tools" subsection

**Agent permissions** · Partly — What an agent may do is fixed at design time by which components you put in Tool Mode and wire to its Tools port; no runtime approval prompt or per-action policy gate is documented inside Langflow itself.

> When in Tool Mode, a component has a Toolset port that you must connect to an Agent component's Tools port if you want to allow the agent to use that component's actions as tools.
> — https://docs.langflow.org/agents · "Use the Agent component in a flow", step 7 explanation of Tool Mode

## Which models it runs on

**Claude** · Yes — Anthropic Claude models work out of the box through the core Language Model component, with a dedicated Anthropic bundle adding provider-specific parameters such as prefill and a custom API URL.

> Basic Anthropic support is available in all Langflow installations through the core Language Model component. This bundle adds an enhanced Anthropic component with additional provider-specific parameters.
> — https://docs.langflow.org/bundles-anthropic · Tip callout at the top of the Anthropic bundle page

**GPT** · Yes — OpenAI GPT models are supported in every installation via the core Language Model component, and the OpenAI bundle adds extra request parameters plus OpenAI embeddings.

> Basic OpenAI support is available in all Langflow installations through the core Language Model component. This bundle adds an enhanced OpenAI component with additional provider-specific parameters.
> — https://docs.langflow.org/bundles-openai · Tip callout at the top of the OpenAI bundle page

**Open models** · Yes — Open-weight models are reachable two ways: host them with Ollama and register that service in Langflow, or point the OpenAI Compatible provider at any compatible inference endpoint you run.

> Configure the OpenAI Compatible provider to point at any OpenAI-compatible endpoint.
> — https://docs.langflow.org/components-models · "My preferred provider or model isn't listed" callout, first bullet

**Model choice** · Yes — You choose the provider and model per component, and the documentation states plainly that the framework does not tie you to particular LLMs or vector stores.

> it doesn't require you to use specific large language models (LLMs) or vector stores
> — https://docs.langflow.org/ · "What is Langflow?" opening paragraph

## Where you use it

**In your editor** · No — The install page enumerates every way to run Langflow — desktop app, Docker, Python package, from source — and none of them is an editor or IDE extension; the authoring surface is Langflow's own visual editor.

> Langflow can be installed in multiple ways. Choose an option below to jump to the steps.
> — https://docs.langflow.org/get-started-installation · Page intro sentence above the four install option cards (Desktop, Docker, Python package, source)

**On the command line** · Yes — A first-class CLI ships with the Python package and manages the server, API keys, migrations and superusers, though it is unavailable in the desktop build.

> The Langflow command line interface is the main interface for managing and running the Langflow server.
> — https://docs.langflow.org/configuration-cli · "Langflow CLI" page, first sentence

**In your pipeline** · Partly — Flows are callable from any pipeline through the documented HTTP API and API-key header, but no pull-request integration, GitHub action or CI-specific surface is documented by the vendor.

> Validates API keys against the LANGFLOW_API_KEY environment variable. Useful for Kubernetes deployments, CI/CD pipelines, or any environment where you want to inject a pre-defined API key without database configuration.
> — https://docs.langflow.org/api-keys-and-authentication · LANGFLOW_API_KEY_SOURCE table, "env" row

**In a browser** · Yes — The visual editor and Playground are a browser application served by the Langflow server you start, reached on port 7860 by default rather than on a vendor-hosted domain.

> To access Langflow, navigate to `http://localhost:7860/`.
> — https://docs.langflow.org/get-started-installation · "Install and run Langflow with Docker", step 3

## Whose machine it runs on

**Self-hosted** · Yes — Self-hosting is the primary delivery model: a Docker image, a pip-installable package and deployment guides for major clouds, all running on infrastructure you control.

> Langflow is completely open source and you can deploy it to all major deployment clouds.
> — https://github.com/langflow-ai/langflow · Repository README, "Deployment" section

**Open source** · Yes — The full application is published under the MIT licence in the vendor's own public repository, and components can be customised in Python because the source is available.

> Source code access lets you customize any component using Python.
> — https://github.com/langflow-ai/langflow · Repository README, "Highlight features" list; MIT licence badge and repository licence field

## What it costs to run

**How it meters** · No — Langflow itself meters nothing — no seats, tokens, credits or tasks are billed by the vendor; your real spend is the infrastructure you run it on plus the model provider keys you supply.

> Langflow is completely open source and you can deploy it to all major deployment clouds.
> — https://github.com/langflow-ai/langflow · Repository README, "Deployment" section

**Free tier** · Yes — No call, contract or card is involved: install the MIT-licensed package with one pip command, or download the desktop app after a contact form, and run it indefinitely.

> uv pip install langflow
> — https://docs.langflow.org/get-started-installation · "Install and run the Langflow OSS Python package", step 3 install command

**API access** · Yes — A documented REST API with /v1 and /v2 endpoints creates, edits and runs flows, and an OpenAPI spec is served from the running instance at /docs.

> You can use the Langflow API for programmatic interactions with Langflow, such as the following: Create and edit flows, including file management for flows.
> — https://docs.langflow.org/api-reference-api-examples · "Get started with the Langflow API", opening list

**MCP server** · Yes — Each Langflow project is exposed as an MCP server whose flows appear to clients as tools, over streamable HTTP with SSE fallback and optional API-key authentication.

> This page describes how to use Langflow as an MCP server that exposes your flows as tools that MCP clients can use when generating responses.
> — https://docs.langflow.org/mcp-server · "Use Langflow as an MCP server", second paragraph

**Bring your own key** · Yes — Bringing your own keys is mandatory rather than optional: you enter each provider's API key in the Model Providers pane or as a global variable, and inference is billed to your own provider account.

> In the API Key field, add your provider's API key. Some providers require additional configuration fields.
> — https://docs.langflow.org/components-models · "Use language model components in flows", step 1.3

## Buying it for a team

**A company can buy it** · Partly — No software licence is sold, so there is nothing to procure for the engine itself; an organisation can only buy an IBM support subscription around its own deployment.

> IBM Elite Support for Langflow is a subscription to the Langflow expertise at IBM.
> — https://docs.langflow.org/luna-for-langflow · IBM Elite Support for Langflow, paragraph 2

**Seat model** · No — Multi-user accounts and roles exist, but no seat price, seat floor or seat ceiling is published anywhere: the software is free to run for as many users as you provision.

> Langflow seeds three built-in system roles that an enforcement plugin uses as a starting point for policy sync. These roles cannot be deleted.
> — https://docs.langflow.org/authorization · System roles, intro sentence above the role table

**Admin controls** · Partly — An admin can cap external users at viewer, editor or admin from a JWT claim and block their API keys, but the docs state this deny-only ceiling is explicitly not full role-based access control.

> The access ceiling is a coarse, deny-only mechanism that maps a JWT claim value to one of three access levels: viewer, editor, or admin. It is not a full RBAC engine.
> — https://docs.langflow.org/external-authentication · "Configure an access ceiling from JWT claims", opening paragraph

**Audit log** · Partly — Langflow records API-key usage counts and last-used timestamps in its database and can export traces to an OTLP backend, but no administrative audit trail of who changed what is documented.

> By default, Langflow tracks API key usage through `total_uses` and `last_used_at` records in your Langflow database.
> — https://docs.langflow.org/api-keys-and-authentication · "Track API key usage" section, first sentence

**Single sign-on** · Partly — OIDC-style single sign-on works by trusting a gateway or identity provider token, validated against a JWKS endpoint with just-in-time user provisioning; no native SAML support is documented.

> External authentication lets an upstream identity provider, OIDC proxy, or corporate SSO gateway handle login. Langflow accepts the token the proxy forwards, validates it against the identity provider's JWKS endpoint
> — https://docs.langflow.org/external-authentication · "External authentication" page, first paragraph

_Not established: Pooled budget_

## What happens to your code

**Opt out of training** · Partly — Anonymous usage telemetry is on by default and switched off with one variable, but the vendor makes no statement about training models on your data either way.

> To opt out of telemetry, set DO_NOT_TRACK=True in your Langflow environment variables before starting Langflow. This disables telemetry data collection.
> — https://docs.langflow.org/contributing-telemetry · Opt out of telemetry, first sentence

**Data residency** · Partly — Residency follows your deployment choice rather than a vendor commitment: flows and chat history live in the database of the server you run, and the vendor publishes no jurisdiction guarantee.

> Langflow is completely open source and you can deploy it to all major deployment clouds.
> — https://github.com/langflow-ai/langflow · Repository README, "Deployment" section

**Getting out** · Yes — Flows export to JSON, optionally stripped of API keys, and the underlying database is a file you control, so there is no lock-in and no credit balance that could expire.

> Additionally, when exporting flows, you can choose to omit API keys from the exported flow JSON.
> — https://docs.langflow.org/api-keys-and-authentication · "Component API keys" section, final paragraph

**Certifications** · No — No SOC 2, ISO 27001 or GDPR attestation is claimed; the single GDPR mention assigns the obligation to whoever operates the server.

> Follow industry best practices and legal requirements for data privacy and encryption of data in transit and at rest, including GDPR requirements, HTTPS, TLS, and SSL.
> — https://docs.langflow.org/deployment-prod-best-practices · Security, Encryption and privacy bullet

## Provenance

Every fact above comes from the vendor. Nothing here is independently corroborated yet.

## Coverage elsewhere

None recorded. Every fact above comes from the vendor and is not independently corroborated yet.

## What changed

- 2026-09-25 — Capabilities · Commercial terms · Free tier · Price
- 2026-09-03 — Capabilities · Commercial terms · Free tier

## Correcting this record

Anything marked "Not established" is a question we have not answered — it is not a No. If you represent Langflow, or anything here is wrong or out of date, you can supply it with a source: the operations this site accepts from an agent are listed at https://theagentsindex.com/.well-known/agents.json — that catalogue needs no key to read, and each entry states its own auth.

Supplying evidence is free. Nothing that can be bought here changes what a field says, whether a gap is closed, or how this record is scored.