Cut LLM Costs with Automatic Provider Fallback

Route one model id across upstreams so AnyRouter picks the cheapest healthy provider and falls back when one fails.

Add a provider block to a Chat Completions request and AnyRouter picks among the upstreams that serve the model, skipping any that are failing. Prices are US dollars per million tokens.

Field	Effect
`sort`	Preference, such as `"price"` (cheapest healthy first)
`order`	Explicit provider sequence
`max_price`	Skip upstreams above the cap
`allow_fallbacks`	Move on when an upstream errors or is rate-limited; `false` fails fast

Cheapest healthy provider first, with fallback:

curl https://anyrouter.dev/api/v1/chat/completions \
  -H "Authorization: Bearer sk-ar-v1-your-key" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "openai/gpt-oss-120b",
    "messages": [{"role": "user", "content": "Summarize this changelog in 3 bullets."}],
    "provider": {
      "sort": "price",
      "max_price": { "prompt": 0.5, "completion": 1.5 },
      "allow_fallbacks": true
    }
  }'

To control the order instead, use "order": ["deepinfra", "cerebras"]: AnyRouter tries those first, then any other healthy upstream when allow_fallbacks is on. If no upstream is under max_price, the request returns an error rather than picking a pricier provider.

Save these preferences once as a preset and use its slug as the model.

AnyRouter routes requests to third-party models. Models and outputs are provided as is, with no warranty; you use them at your own risk. Disclaimer