Dify
Open-source canvas for LLM workflows, RAG knowledge bases and agents
Overview
Dify is an open-source platform from LangGenius, Inc. for building LLM applications on a visual canvas: workflows with human review steps and triggers, RAG knowledge pipelines and sandboxed agents, each published as an API, web app or MCP server. It can be self-hosted under a modified Apache 2.0 licence or used on Dify Cloud.
Pricing
Pricing shown for reference only. These figures reflect RECATOOLS research as of 14 Oct 2026 and may be out of date or incomplete. This is not financial or purchasing advice — always confirm the current price on the provider’s official website before making any decision.
- All core features in the public repository
- Single workspace
- 200 message credits
- 1 team member, 5 apps
- 50 knowledge documents, 50MB storage
- 30 days of logs
- 5,000 message credits a month
- 3 team members, 50 apps
- 500 knowledge documents, 5GB storage
- 20,000 trigger events a month
- 10,000 message credits a month
- 50 team members, 200 apps
- 1,000 knowledge documents, 20GB storage
- Unlimited trigger events
- Commercial licence authorisation
- Multiple workspaces and SSO
- Negotiated SLAs and official support
What you can produce with Dify
- A RAG chatbot over your own documents, built from a knowledge pipeline without writing retrieval code
- Multi-step workflows with branching, loops and a human-input node that pauses for review
- Workflows started automatically by a schedule, a webhook or a plugin event
- Any Dify app published as an API, an embeddable web app or an MCP server
- A self-hosted deployment on Docker or Kubernetes with your own model provider keys
- A sandboxed agent that runs commands and works with files, used alone or inside a workflow (beta)
Dify gives you a visual workflow builder, a RAG pipeline and agents in one self-hostable package, and everything you build becomes an API or MCP server straight away. Releases are frequent, and the August 2026 sandboxed Agent shows it moving beyond chat apps. The free Sandbox is a trial, not a plan: 200 message credits in total and one member.
Two cautions. The licence bars running a multi-tenant service from the source without permission, so check it before building a SaaS on Dify. And 13 security advisories in 2026, several cross-tenant, mean self-hosters must patch promptly. For internal AI apps you want to own, it is an excellent starting point.
What we found
Dify is an open-source platform for building LLM applications on a visual canvas: chatbots, workflows, retrieval-augmented knowledge bases and agents, each published as an API, a web app or an MCP server. It comes from LangGenius, Inc., a Delaware company that a March 2026 funding release placed in Menlo Park and that its chief executive, Luyu Zhang, leads. The project went public on GitHub in April 2023 and ships often; version 1.17.0 in August 2026 added a beta Agent that works inside its own sandbox, workspace-level skills, E2B cloud sandboxes and human approval steps inside loops.
The workflow side is where it is strongest. Nodes cover LLM calls, knowledge retrieval, code, HTTP requests, branching, iteration and a human-input node that pauses a run for review, and workflows can start on a schedule, a webhook or a plugin event. A knowledge pipeline handles ingestion and chunking, and a marketplace supplies model providers and tools.
Read the licence before you plan anything commercial. It is Apache 2.0 with conditions: you may not run a multi-tenant service from the source without written permission, and you may not remove the logo or copyright from the frontend. GitHub classifies it as "Other". Security also deserves attention: 13 advisories were published against the project in 2026, several of them cross-tenant data exposures, so self-hosters need to keep up with releases.
Dify Cloud starts with a free Sandbox: 200 message credits in total, one member, five apps and 50 knowledge documents. Professional costs $59 a month or $590 a year, Team $159 or $1,590, and Enterprise is quoted with SSO and a commercial licence. In March 2026 LangGenius raised a $30 million Series Pre-A led by HSG at a $180 million valuation, and it says Dify runs on more than 1.4 million machines, with 280 enterprise customers including Maersk and Novartis. Both figures are the company's own.
Compiled by RECATOOLS from public sources: vendor documentation and pricing, code repositories and published reporting.
About this listing
This entry was compiled from publicly available data including Dify's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with Dify unless explicitly stated.
Third-party AI tools update their pricing, features, availability, and policies frequently. Information here may be outdated by the time you read this — we make reasonable efforts to keep listings current, but cannot guarantee absolute accuracy.
For the latest details, please refer to Dify directly →
Spotted something out of date? Suggest an update →
Related reading
From our guides
Guide · 16 min read
Self-Host Dify with Docker Compose: 15 Containers to Your Own LLM App Studio (2026)
We deployed Dify 1.16.0 on Docker Desktop, hit a real port clash, wired DeepSeek...
Guide · 13 min read
n8n vs Dify vs Langflow vs Flowise: Which Visual AI Workflow Builder Fits You? (2026)
Four self-hostable visual AI workflow builders, four very different licenses and...
Alternatives to Dify
More in Agents & Automation
