LLMRouter supports 16+ routing models across single-round, multi-round, agentic, and personalized categories. Morph's Model Router is a hosted single-round router worth adding as a baseline: a classifier trained on millions of coding prompts that returns difficulty, ambiguity, and domain in ~50ms, then maps to a model via policy. Useful as a hosted comparison point against the trained KNN/SVM/MLP routers already in the library.
One POST, OpenAI-compatible (/router/multimodel takes allowed_providers + policy, returns a single model name):
import requests
model = requests.post(
"https://api.morphllm.com/v1/router/multimodel",
headers={"Authorization": f"Bearer {MORPH_API_KEY}"},
json={"input": prompt, "allowed_providers": ["openai"], "policy": "balanced"},
).json()["model"]Docs: https://docs.morphllm.com/sdk/components/router
LLMRouter supports 16+ routing models across single-round, multi-round, agentic, and personalized categories. Morph's Model Router is a hosted single-round router worth adding as a baseline: a classifier trained on millions of coding prompts that returns difficulty, ambiguity, and domain in ~50ms, then maps to a model via policy. Useful as a hosted comparison point against the trained KNN/SVM/MLP routers already in the library.
One POST, OpenAI-compatible (
/router/multimodeltakesallowed_providers+policy, returns a single model name):Docs: https://docs.morphllm.com/sdk/components/router