What is the best OpenAPI mock server?

It depends on your runtime. APIGhost is a Python CLI that generates a mock from a spec and adds VCR record/replay and named scenarios. Stoplight Prism is the mature Node option for spec-driven mocking and validation. WireMock is the JVM choice for heavyweight stubbing and record-and-playback. Mockoon offers a visual editor plus a CLI.

Does Prism support record and replay?

Prism works as a mock and as a validation proxy, but it does not provide VCR-style cassette record/replay. Example selection is done via the HTTP Prefer header rather than saved scenario presets.

OpenAPI Mock Server Comparison:
APIGhost vs Prism vs WireMock vs Mockoon

Q: Does Mockoon have a command-line interface?

Yes. In addition to its desktop GUI, Mockoon ships an official CLI (@mockoon/cli), a Docker image, and a GitHub Action, so it can run mock APIs headless in CI.

Four solid ways to stand up a mock API from an OpenAPI spec. They overlap a lot, but each has a different sweet spot. Here is an honest, side-by-side look so you can pick the right one for your stack.

Last reviewed: June 2026. Tool capabilities change — links to each project's docs are included so you can verify the current state.

Quick pick

You are a Python team and want one CLI that does it all. Use APIGhost — generate a mock from your spec, plus VCR-style record/replay and named edge-case scenarios, installed with pip.

You want the most mature spec-driven mock + a validation proxy (Node). Use Prism by Stoplight.

You need heavyweight stubbing, fault injection, and record-and-playback on the JVM. Use WireMock.

You want a visual editor to design and tweak mocks by hand. Use Mockoon (it also has a CLI and Docker image).

Feature comparison

Capability	APIGhost	Prism	WireMock	Mockoon
Mock generated directly from an OpenAPI spec (no stub-writing)	Yes	Yes	Partial	Yes
Schema-aware fake data	Yes	Yes	Partial	Yes
VCR-style record & replay (deterministic cassettes)	Yes	No	Yes	Partial
Named scenario presets for edge cases	Yes	Partial	Yes	Yes
Runs headless from the CLI	Yes	Yes	Yes	Yes
Visual editor / GUI	No	No	Cloud	Yes
Runtime & install	Python · `pip`	Node · `npm`	JVM · jar / Docker	Node · `npm` / app

Yes = built-in | Partial = available via an add-on, proxy mode, or hosted tier | No = not supported. "Partial" notes are explained per tool below.

APIGhost

APIGhost is a Python CLI that turns an OpenAPI 3.0/3.1 spec into a running mock server in one command, populates it with Faker-powered fake data keyed off property names, and adds two things most spec-driven mocks don't bundle: VCR-style record/replay cassettes for deterministic test runs, and named scenario presets for exercising edge cases like error responses.

pip install --index-url https://coding-dev-tools.github.io/pypi-index/simple/ apighost
# or, install straight from source:
# pip install git+https://github.com/Coding-Dev-Tools/apighost.git
apighost serve petstore.yaml          # mock from a spec
apighost serve spec.yaml --scenario error
apighost record petstore.yaml          # capture a deterministic cassette
apighost replay my-recording           # replay it forever

Best fit: Python-centric teams who want spec-driven mocking and deterministic replay and scenario switching from a single pip install, without bringing in a Node toolchain or the JVM. Free for local/dev use; a Pro tier ($12/mo) adds unlimited requests, VCR cassettes, and CI/CD integration. Repo & docs →

Stoplight Prism

Prism is a mature, open-source Node CLI that turns OpenAPI v2/v3 (and Postman) documents into a mock server. Run it with the -d flag and it generates dynamic, schema-valid data via Faker.js. It also works as a validation proxy in front of a real API. You can steer which example or status code it returns using the HTTP Prefer header (for example Prefer: example=... or Prefer: code=...) — that is the "Partial" mark in the scenarios row: it is example selection rather than saved, named presets. Prism has no VCR-style cassette record/replay.

Best fit: Node teams who want the most established spec-driven mock plus request validation. Project page →

WireMock

WireMock is a heavyweight, battle-tested API mocking tool on the JVM. Its core strengths are rich request matching, stateful Scenarios, fault and latency injection, and a genuine record-and-playback mode (proxy an existing API, then snapshot the traffic into stub mappings). The "Partial" mark on spec-driven generation reflects that WireMock's OSS core is oriented around stub mappings and proxy recording; auto-generating stubs straight from an OpenAPI document is strongest in WireMock Cloud and extensions rather than the base CLI. Its fake-data story is Handlebars templating with helpers rather than schema-derived generation.

Best fit: JVM shops, or anyone who needs the richest stubbing, stateful behavior, and resilience testing. Project site →

Mockoon

Mockoon is best known for its friendly desktop GUI for designing mock APIs by hand, but it also ships an official CLI (@mockoon/cli), a Docker image, and a GitHub Action, so it runs headless in CI too. It can import an OpenAPI document, supports templating with Faker helpers, route response rules, and a proxy mode (the "Partial" record mark — proxy mode can capture and log calls rather than producing VCR cassettes).

Best fit: teams who want to visually design and tweak mocks, then export/run them from the CLI. Project site →

So which should you use?

If your stack is Python and you want spec-driven mocking, deterministic record/replay, and named scenarios in one lightweight CLI, APIGhost is the most direct fit. Prism is the closest Node-based alternative for pure spec-driven mocking and validation. WireMock is the right call when you need heavyweight stubbing and record-and-playback on the JVM. Mockoon wins when a visual editor matters. All four are honest, capable tools — the best choice is mostly about your runtime and whether record/replay and scenarios are part of your workflow.

Try APIGhost on GitHub

← Back to the DevForge toolkit

OpenAPI Mock Server Comparison:APIGhost vs Prism vs WireMock vs Mockoon