Most teams wire a provider client, then discover they need a second model. The refactor is never one line.
Hisoka keeps the OpenAI chat-completions shape and moves the model into the request body:
resp = client.chat.completions.create(
model="another-model-id",
messages=[{"role": "user", "content": "Hi"}],
)The base URL, the key, and the rest of the request stay untouched. Switching model becomes a string change, not a migration.
