Inworld Router

One endpoint. The right model for every user and context.

A single endpoint that routes every request to the right model for that user. Optimize for cost, latency, engagement, revenue, or any metric you care about. 220+ models, no markup on provider rates.

Explore Models Read the docs

curl 'https://api.inworld.ai/v1/chat/completions' \
  -H "Content-Type: application/json" \
  -H "Authorization: Basic $INWORLD_API_KEY" \
  -d '{

    "model": "inworld/user-aware",

    "messages": [{"role": "user", "content": "Hello"}],
    "extra_body": {
        "metadata": {
            "language": "es",
            "country": "MX",
            "plan": "free"
        }
    }
  }'

curl 'https://api.inworld.ai/v1/chat/completions' \
  -H "Content-Type: application/json" \
  -H "Authorization: Basic $INWORLD_API_KEY" \
  -d '{

    "model": "inworld/context-aware",

    "messages": [
        {"role": "system", "content": "You are a support agent."},
        {"role": "user", "content": "My order never arrived"}
    ],
    "extra_body": {
        "metadata": {
            "intent": "complaint",
            "emotion": "frustrated",
            "session_turns": 4
        }
    }
  }'

220+

Models

Markup

Line to integrate

See it in action

Try each routing strategy live.

Ready to build? Get your API key in under a minute.

The router that knows your users

One integration. Route by who your user is, optimize automatically, run real experiments, and never go down. No infrastructure to manage.

Route by who the user is

Pass any metadata with each request and route to a different model based on language, region, subscription tier, or any signal you define. Each user gets the model that fits them best.

Optimize for any metric you care about

Route to the cheapest model, the fastest, or the one that drives the strongest engagement or revenue. Set your optimization target per route and let the router execute.

A/B test models against real outcomes

Split traffic by weight with sticky user assignment. Ramp a new model to 5% of requests, watch the metrics, then dial it up. Find the winner.

Maximize reliability, minimize downtime

When a provider returns an error, the router retries the next model in your fallback chain instantly. No retry logic to write. Your users never notice.

Drop-in OpenAI compatibility

Change one line: swap your base URL and model string for an Inworld router ID. Every OpenAI-compatible SDK and client works immediately. No migration, no rewrite.

Per-request observability

Every request logs the model selected, TTFT, cost per token, and the full attempt chain including any failovers. Push to your analytics platform of choice.

Every routing use case, one endpoint

Common patterns for Inworld Router in production. Each one is a real configuration you can copy, adapt, and deploy.

curl 'https://api.inworld.ai/v1/chat/completions' \
  -H "Authorization: Basic $INWORLD_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "inworld/user-aware",
    "messages": [{"role": "user", "content": "Hola, necesito ayuda"}],
    "extra_body": {
      "metadata": {
        "language": "es",
        "country": "MX",
        "plan": "free"
      }
    }
  }'

curl 'https://api.inworld.ai/router/v1/routers' \
  -H "Authorization: Basic $INWORLD_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "displayName": "User-Aware",
    "routes": [
      {
        "condition": {
          "cel_expression": "language == \"es\""
        },
        "route": {
          "variants": [
            { "variant": { "modelId": "openai/gpt-5.2" }, "weight": 100 }
          ]
        }
      },
      {
        "condition": {
          "cel_expression": "plan == \"free\""
        },
        "route": {
          "variants": [
            { "variant": { "modelId": "anthropic/claude-haiku-4-5" }, "weight": 100 }
          ]
        }
      }
    ],
    "defaultRoute": {
      "variants": [
        { "variant": { "modelId": "anthropic/claude-sonnet-4-6" }, "weight": 100 }
      ]
    }
  }'

curl 'https://api.inworld.ai/v1/chat/completions' \
  -H "Authorization: Basic $INWORLD_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "inworld/user-aware",
    "messages": [{"role": "user", "content": "Hola, necesito ayuda"}],
    "extra_body": {
      "metadata": {
        "language": "es",
        "country": "MX",
        "plan": "free"
      }
    }
  }'

curl 'https://api.inworld.ai/router/v1/routers' \
  -H "Authorization: Basic $INWORLD_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "displayName": "User-Aware",
    "routes": [
      {
        "condition": {
          "cel_expression": "language == \"es\""
        },
        "route": {
          "variants": [
            { "variant": { "modelId": "openai/gpt-5.2" }, "weight": 100 }
          ]
        }
      },
      {
        "condition": {
          "cel_expression": "plan == \"free\""
        },
        "route": {
          "variants": [
            { "variant": { "modelId": "anthropic/claude-haiku-4-5" }, "weight": 100 }
          ]
        }
      }
    ],
    "defaultRoute": {
      "variants": [
        { "variant": { "modelId": "anthropic/claude-sonnet-4-6" }, "weight": 100 }
      ]
    }
  }'

FAQ

Yes. Migration guides are available for OpenRouter and Anthropic-based setups. The core change is updating your base_url and API key, while your existing request structure stays the same.

Router provides access to hundreds of models from leading providers, such as OpenAI, Anthropic, Google, and many more. You can see the full model list here.

While Router is in Research Preview, you pay provider rates directly, with no markup or margin added. Rates for all models are available here.

Inworld Router itself doesn't impose additional rate limits on top of providers. Provider-level rate limits are handled automatically by retrying the next model in your fallback chain.

Most gateways give you a unified API and basic fallback. Inworld Router offers more control and lets you run real experiments: conditional routing, dynamic tiering, traffic splitting by percentage, sticky user assignment, with results pushed to your analytics platform of choice.