Use case · LLM routing

Route every prompt to the right model

Send the user turn and your route options; get back one choice and a probability for each option. Your code decides what to do with it: send easy prompts to a cheaper model, hard ones to a stronger one.

Get API key Read the quickstart

Example request

One call, five lines

Set DECISIONMODELS_API_KEY to a key from your dashboard. The request below uses the EU tier; send S1-Region: global for the Peer-to-Peer tier.

Request · POST /v1/systemone

curl https://api.decisionmodels.io/v1/systemone \
  -H "Authorization: Bearer $DECISIONMODELS_API_KEY" \
  -H "Content-Type: application/json" \
  -H "S1-Region: eu" \
  -d '{"model":"s1-fast","state":"Summarise this 40-page contract and flag unusual liability clauses.","questions":{"route":{"type":"choice","instructions":"Which model tier should answer this prompt?","criteria":{"small_model":null,"large_model":null,"needs_human":null}}}}'

Example response

A typed answer with probabilities

Illustrative, not a measured model output. Always read the probability before your application acts on it.

Response · illustrative

{"id":"dec_example","model":"s1-fast","answers":{"route":{"type":"choice","choice":"large_model","confidence":0.82,"probabilities":{"small_model":0.11,"large_model":0.82,"needs_human":0.07}}},"usage":{"input_tokens":214,"output_tokens":0,"decisions":1},"tier":"eu"}

Notes

Before you build on it

The routes are yours: name them after your own models or queues.
Output tokens are not billed, so a decision costs only the input you send.
For a ready-made classifier that returns category, difficulty, stakes and routing flags in one call, see s1-llm-auto-router.