I built a mock API tool because I was tired of frontend work blocking on a backend that didn't exist yet. The design choice that mattered most: the data is seeded, not random.
Pick a dataset (say, users) and a seed, and you get exactly the same rows every time: same names, same emails, same IDs. That matters because:
- Tests that assert on specific values don't break on every run
- Snapshot tests of rendered output stay stable
- Changing the seed gives you a different but equally consistent dataset
On top of that is a real in-memory router, not canned responses. GET /api/mock/users?country=Germany&sort=-createdAt really filters and sorts the rows, with page/limit/cursor pagination, search, field selection and nested relations like /users/:id/posts.
Other pieces: mock JWT/API-key/basic auth, error simulation (400-503 or random) and latency presets for testing loading and failure states, GraphQL schema browsing, OpenAPI generation, code snippets in 15 languages, and mock WebSocket/SSE streams.
The part that trips people up: there are two modes. Simulate runs entirely in your browser (offline, zero network calls). Live endpoint hits a hosted URL with CORS enabled, which you need if a separate app or a teammate must call it.
Try: curl "https://forgeplug.com/api/mock/users?limit=5"
Tool: https://www.forgeplug.com/tools/mock-api-studio
Do you seed your mock data, or generate it fresh per run? I'd like to hear what breaks in your tests.
![Multi Robot and Swarm Physical AI — Ch.00: Introduction [EN]](https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Farticles%2F3kjs92ynywy72p5gq7o2.png)

