Quickstart

Model Router speaks the OpenAI Chat Completions API. Point your existing SDK at our base URL, set model: "auto", and you're routing.

1. Drop in the base URL

Python:

python

from openai import OpenAI

client = OpenAI(
    base_url="https://router.aiskillhub.info/api/v1",
    api_key="mr_live_...",
)

res = client.chat.completions.create(
    model="auto",  # router picks the cheapest capable model
    messages=[{"role": "user", "content": "Classify this ticket sentiment"}],
)
print(res.choices[0].message.content)
# Response headers tell you what happened:
#   x-router-model:     deepseek/deepseek-v3
#   x-router-tier:      simple
#   x-router-saved-pct: 94.4

2. Or just cURL it

bash

curl https://router.aiskillhub.info/api/v1/chat/completions \
  -H "Authorization: Bearer $MODEL_ROUTER_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "auto",
    "messages": [{"role": "user", "content": "Write a haiku about cheap tokens"}]
  }'

3. Bring your own keys

Add your provider keys (OpenAI, Anthropic, DeepSeek, Google, MiMo…) in the Dashboard. We encrypt them at rest (AES-256-GCM) and only use them to forward routed requests. Or set a single Vercel AI Gateway key server-side to reach every provider at once.

4. Read the routing headers

x-router-model — which model handled it
x-router-tier — classified difficulty
x-router-saved-usd / x-router-saved-pct — savings vs GPT-4o baseline
x-router-fallback — set if a fallback model was used

Try it locally right now

bash

# Local demo key (seeded): mr_live_demo_key_for_local_testing_001
curl http://localhost:3000/api/v1/chat/completions \
  -H "Authorization: Bearer mr_live_demo_key_for_local_testing_001" \
  -H "Content-Type: application/json" \
  -d '{"model":"auto","messages":[{"role":"user","content":"hi"}]}'