Freemiumautomation

Dify

Build production-ready agentic workflows in the cloud or on your own infrastructure.

A collaborative platform for building and deploying agentic workflows, RAG pipelines, and AI applications.

4.4

Try Dify Free →

Opens official website

AI agentsworkflow builderRAGself-hostedLLM apps

Live change feed

Recent verified changes

Shared with the homepage and tool pages

difyAug 25, 2026

Dify 1.17.0 adds cloud agent sandboxes, snapshots, and workspace skills

Dify 1.17.0 adds an E2B cloud-sandbox backend, published-agent home-directory snapshots, and workspace-level skill management with draft, publish, and version lifecycle. Self-hosted teams should review the new runtime and access configuration before upgrading.

Verify official source →

difyJul 28, 2026

Dify 1.16.1 hardens agent sandboxes and workflow security

Dify 1.16.1 isolates the local agent sandbox behind a restricted proxy, adds token authentication between internal services, blocks template-injection risks, and patches a React Server Components denial-of-service issue.

Verify official source →

verifiedLast verified 2026-08-31

Dify 1.17.0 agent runtime capabilities verified

Dify 1.17.0 adds an E2B cloud-sandbox option, published-agent home-directory snapshots, and workspace-level skills with draft, publish, and version lifecycle. Managed cloud, Community self-hosting, and Enterprise deployment remain available; self-hosted teams should review new runtime and access configuration before upgrading.

Agent runtime: Local sandbox or E2B cloud-sandbox backend
Published agents: Home-directory snapshots can restore prepared agent state
Skills: Workspace draft → publish → version lifecycle
Sandbox: Free with limited apps, knowledge, credits, and logs
Professional: $59 per workspace/month; $590 annually
Team: $159 per workspace/month; $1,590 annually
Deployment: Managed cloud, Community self-hosting, or Enterprise

Verify source: Dify 1.17.0 official release notes →

About Dify

Dify is a visual platform for building, testing, deploying, and monitoring AI applications. It brings workflows, chatflows, agents, retrieval-augmented generation (RAG), model-provider configuration, plugins, APIs, logs, and team workspaces into one product. Teams can start with the managed Dify Cloud service, self-host the Community edition, or discuss private Enterprise deployment. Dify is most useful when a team wants application logic to remain visible and editable instead of burying prompts, retrieval, and tool calls inside custom code.

✅ Pros

+Visual workflow and chatflow builders make multi-step application logic easier to inspect
+Combines agents, RAG knowledge bases, model providers, plugins, APIs, and monitoring
+Supports managed cloud, free Community self-hosting, and private Enterprise deployment paths
+Can publish the same application as a web app or API
+Lets teams bring their own model-provider API keys after included message credits are exhausted
+Official templates and a growing plugin marketplace reduce setup work

❌ Cons

−Production deployments still require testing, evaluation, access controls, and operational monitoring
−Cloud quotas vary by plan across apps, documents, storage, trigger events, members, and knowledge requests
−Self-hosting transfers upgrades, backups, security, model costs, and infrastructure operations to the user
−The number of concepts—workflows, chatflows, agents, datasets, plugins, and model providers—creates a learning curve
−Included message credits are not the same as unlimited model usage; external model API charges may still apply

Key Features

Workflow and Chatflow Studio

Connect model, retrieval, logic, tool, code, and output nodes in a visual execution path that teams can test and revise.

Agents and Tool Use

Build assistants that can select approved tools and complete multi-step tasks within configured instructions and limits.

Knowledge Pipelines and RAG

Import, process, index, retrieve, and manage documents or external knowledge sources for grounded applications.

Models and Plugins

Configure multiple model providers and extend applications with marketplace or custom tools, data sources, models, and triggers.

Deploy and Observe

Publish applications as web experiences or APIs, then use logs, feedback, annotations, and integrations to review runtime behavior.

Cloud or Self-Hosted

Choose a managed workspace for speed, the Community edition for infrastructure control, or an Enterprise deployment path.

Pricing Plans

Dify Cloud lists a free Sandbox plan, Professional at $59 per workspace/month or $590 annually, and Team at $159 per workspace/month or $1,590 annually. The Community edition is free to self-host, but infrastructure and model-provider costs remain separate. Verified July 29, 2026; quotas and prices may change.

🏆 Our Verdict

Dify is a strong choice for teams that want to move from an AI application idea to a visible, deployable workflow without building every orchestration component from scratch. Its biggest advantage is breadth: workflows, agents, RAG, model access, plugins, deployment, and monitoring live in one environment. That breadth is also the main tradeoff. Buyers should compare the convenience of Dify Cloud with the operational responsibility of self-hosting, then estimate model, storage, trigger, and team usage rather than choosing only by the headline workspace price.

✅ Best For

Teams building internal AI assistants and knowledge applicationsDevelopers and product teams prototyping LLM applications before writing a custom orchestration layerOperations teams that need visible, repeatable AI workflowsOrganizations evaluating cloud and self-hosted deployment optionsBuilders combining multiple model providers, RAG, tools, and APIs

⚠️ Not Recommended For

Users who only need a general-purpose hosted chatbotTeams unwilling to test and monitor production AI workflowsSelf-hosters without a plan for upgrades, backups, security, and model costsBuyers who need a fixed all-inclusive price for model usage

💬 Frequently Asked Questions

What is Dify?

Dify is a platform for building AI applications with visual workflows, chatflows, agents, RAG knowledge bases, model providers, plugins, APIs, and runtime monitoring.

Is Dify free?

Dify Cloud has a free Sandbox plan with usage limits. The Community edition can also be self-hosted without a Dify subscription, although infrastructure, operations, and model-provider usage still have their own costs.

Does Dify include AI model usage?

Dify Cloud includes a limited number of message credits that vary by plan and model. When those credits are exhausted, users can connect their own model-provider API keys and pay the applicable provider charges.

Should I use Dify Cloud or self-host Dify?

Choose Dify Cloud when managed setup and faster onboarding matter most. Consider self-hosting when infrastructure control is important and your team can handle upgrades, backups, security, monitoring, and scaling.

Who should use Dify?

Dify is best suited to product, developer, operations, and knowledge teams building repeatable AI applications that combine models, retrieval, tools, APIs, and visible workflow logic.

Can Dify applications be published for other users?

Yes. Dify documents publishing applications as web apps and APIs. Available limits, branding, collaboration, and operational controls depend on the selected plan and deployment model.