Dify is a source-available platform for visually building LLM chat apps, agents and workflows with knowledge bases, Marketplace model and tool plugins, human-input pause nodes, run cloud-hosted or self-hosted.
Human Input web-app delivery not available in Trigger-started workflows; request closes after first response; default 3-day timeout; upload caps (e.g. 15 MB documents); Sandbox log history 30 days
plan limits from https://dify.ai/pricing/dify-cloud
Fit and decision conditions
Prerequisites
Community Edition: self-host with Docker; Enterprise: Helm chart / Terraform modules for AWS, GCP, Azure; Cloud: message credits per plan, then own model API key
Sandbox free (200 message credits); Professional $59/workspace/month or $590/year (3 members, 5,000 credits/month); Team $159/workspace/month or $1,590/year (50 members, 10,000 credits/month); taxes extra
Dify Cloud region not named ("in our managed cloud region"); Enterprise: Dify-managed single-tenant cluster "in your region" or customer-managed in own AWS/GCP/Azure account; privacy policy lists processing in US, Australia, Canada, China, EEA and UK
Cloud wording from https://dify.ai/pricing/dify-cloud; processing locations from https://dify.ai/legal/privacy-policy (website servers 'located in the United States'). No named EU region for Dify Cloud
Does it train AI on your data?
Not used for training: "We do not train AI models ourselves, and we will not use your AI interaction data for model training."
With own API key, training depends on customer's agreement with that model provider; sub-processors 'prohibited from using your data for model training'
Certificates and audit reports
SOC 2 Type II (attestation, auditor Sensiba); ISO 27001:2022 (assessed by Johanson); scope boundary not stated
The provider’s own sources say different things here. The trust centre names SOC 2 Type 1; the March 2026 blog and enterprise page name Type II. Confirm the applicable report.
Named: AWS, Google Workspace, Stripe, GitHub/Google SSO, Metabase, Cloudflare; bring-your-own model via own API key; message credits cover OpenAI and other models; air-gapped deployments can "connect locally deployed models"
Credits and API keys follow the pricing page; local-model support follows the enterprise page.
Retention and deletion
Sandbox logs 30 days then "permanently deleted"; Professional/Team conversation logs kept for active subscription, workflow run logs ~3 months then monthly archives; account information kept while active, then deleted or anonymised
account retention from https://dify.ai/legal/privacy-policy; deleting logs does not delete uploaded files
Export
Archived workflow run logs downloadable monthly as ZIP of CSVs by owner/admins; GDPR right to receive personal information "in a usable electronic format"
Human Input node: "Pause workflows to request human input"; recipients choose a decision button that routes the workflow; delivery via web app or email; 3-day default timeout branch
Logs of conversations and workflow runs incl. raw prompt, "Agent tool iterations" and node-by-node traces; Enterprise: "Tamper-evident audit logs streamed to your SIEM"
SIEM audit logs from https://dify.ai/dify-enterprise
Questions for the provider
Ask the provider for applicable certificates and attestations, including their scope and validity.
Cost per accepted outcome
Editable hypothetical example · not provider pricing or measured performance. Use this monthly worksheet to test your own quote and quality assumptions.
/ accepted outcome
accepted outcomes / month
5,000 tasks × 80% = 4,000 accepted outcomes per month. Higher entered costs raise unit cost; more accepted outcomes spread the same cost over more results.
Formula: entered monthly costs ÷ (monthly tasks × acceptance % ÷ 100). The numerator includes only the costs you enter. Add applicable licence, usage, integration, human review and rework costs yourself; this is not automatically a fully loaded estimate or a vendor billing model. Capacity released is not realized financial value.
Inputs stay in this browser tab’s session, including reload and language changes. They are not added to the page URL. Continuing prepares an unsent enquiry draft.
A written comparison of up to three AI offerings · €195 net
You get the facts that matter for your decision, each with its source, the questions to ask each vendor and a cost estimate based on your own numbers. First you agree your three most important criteria with Sten Franke.
We deliver your brief by email within a maximum of five working days after payment is confirmed and we have agreed the question and minimum information needed to begin. ETHORITY confirms the start date and delivery date and delivers earlier when the reviewed brief is ready. Missing optional information is recorded and assessed; it does not require you to research the solution yourself. Review before delivery and email support are included.
ETHORITY®