Mannequin

Mannequin is an OpenAI-compatible MITM proxy for testing AI-enabled
applications. It sits between your application and a real backend model
API, records every request/response exchange to disk, and later replays
those exchanges without calling the real backend, so your tests and demos
run fast, offline, and for free.

mannequin record proxies live traffic to a backend such as https://api.openai.com/v1 while saving every exchange; point your
application at http://127.0.0.1:8080 instead of the real API. Then
mannequin replay serves the recorded responses back with no API key and
no network calls, so test suites, CI pipelines, and demos run
deterministically at zero cost.