Live router demo
Try the router
Paste a prompt. We'll show you which model we'd send it to and why — including cost, quality score, and median latency on our network.
Your prompt208/8000
The router picked
Click Resolve to see the routing decision.
Routing breakdown
What else we considered
Every candidate the router scored before locking in. The picked row is highlighted.
Resolve a prompt to see the candidate breakdown.
Drop in
Use it in your code
Same OpenAI-compatible API. Swap the base URL and pass orcarouter/auto.
from openai import OpenAI
client = OpenAI(
base_url="https://api.orcarouter.ai/v1",
api_key=os.environ["ORCAROUTER_API_KEY"],
)
resp = client.chat.completions.create(
model="orcarouter/auto",
messages=[{"role": "user", "content": "Summarize this in 3 bullet points: large language model APIs vary in cost and latency, and routing requests to the cheapest acceptable model can save 80%+ on inference spend without changing application code."}],
)
print(resp.choices[0].message.content)Get $5 free credit
Sign up to get an API key and route real traffic. No card required.
