generated: '2026-09-06'
method: searched
source: >-
  https://docs.dify.ai/en/api-reference/guides/get-started,
  https://docs.dify.ai/en/cloud/use-dify/nodes/trigger/webhook-trigger,
  https://dify.ai/pricing
description: >-
  Dify publishes no test-mode credentials and no fixture data. There are no test/live key prefixes,
  no magic identifiers, and no time simulation. What it does publish is an interactive API console
  on every reference page, a separate test URL for workflow webhook triggers, and a free-forever
  self-hosted edition that is the real test environment for most teams.
test_mode:
  present: false
  key_prefixes: null
  note: >-
    Checked, nothing to record. App and knowledge-base API keys have no documented test/live
    distinction and no published prefix convention; the same key hits the same production workspace.
test_fixtures:
  present: false
  note: No test cards, magic identifiers, sample accounts or seeded objects are published.
time_simulation:
  present: false
console:
  present: true
  type: interactive API playground
  location: every operation page under https://docs.dify.ai/en/api-reference/
  description: >-
    Each reference page carries a "Try it" panel. The published quickstart walks a first call through
    it — open Get App Info, click Try it, paste the key into the Authorization field, press Send —
    and states it works for every app type.
  auth_required: true
  docs: https://docs.dify.ai/en/api-reference/applications/get-app-info
webhook_testing:
  present: true
  description: >-
    A workflow's Webhook Trigger exposes a separate test webhook URL alongside its production URL.
    The docs instruct users to always use the test URL when testing, to keep test data out of
    production data. An unpublished trigger must first be put into a listening state with "Run this
    step" or a test run before it will capture a request.
  docs: https://docs.dify.ai/en/cloud/use-dify/nodes/trigger/webhook-trigger
single_node_testing:
  present: true
  description: >-
    Workflows can be debugged one node at a time ("Single Node" / Run this step) with a variable
    inspector and run history, without executing the whole graph.
  docs: https://docs.dify.ai/en/cloud/use-dify/debug/step-run.md
free_tier_as_sandbox:
  plan: Sandbox
  price: free
  quotas: 200 message credits, 5 apps, 50 knowledge documents, 5,000 API requests per month
  note: >-
    Dify's free tier is literally named "Sandbox", but it is a production workspace with small
    quotas — not an isolated test environment. Do not read the plan name as a test mode.
  docs: https://dify.ai/pricing
self_host_as_sandbox:
  present: true
  edition: Community Edition
  description: >-
    The open-source distribution deploys with Docker and runs the same Service API against a local
    base URL, which is the closest thing Dify offers to a disposable test environment.
  docs: https://docs.dify.ai/en/self-host