Reasoning effort

Available reasoning levels depend on the selected provider and API. Different providers of the same model may accept different values.

Check support

Check the selected provider in the model catalog. The /models field endpoints[].supported_reasoning_efforts describes API capabilities and may aggregate several providers; check your selected provider separately. A model name does not guarantee a particular effort level.

The examples use high. Select a model and provider that explicitly support that value.

Configure your API

curl

curl https://api.closerouter.dev/v1/chat/completions \
  -H "Authorization: Bearer $CLOSEROUTER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "YOUR_MODEL_ID",
  "provider": "YOUR_PROVIDER_ID",
  "messages": [
    {
      "role": "user",
      "content": "Explain why the solution works."
    }
  ],
  "reasoning_effort": "high"
}'

Default and disabling

Omitting the parameter uses the model/provider default. The value none explicitly disables reasoning only where supported. Omission and none are not equivalent.