OpenClaw · Laras Router Docs

This guide connects OpenClaw to Laras Router. Every step was checked against the OpenClaw documentation.

You need an API key (lr_live_…) and a topped-up balance. New to Laras Router? Start with the Quickstart.

Set up

Set your API key

Make your key available to the OpenClaw Gateway process.

Shell

export LARAS_API_KEY="lr_live_your_key_here"

Add the provider

In ~/.openclaw/openclaw.json, add a provider under models.providers with api set to openai-completions, and make its model the default.

openclaw.json

{
  agents: {
    defaults: {
      model: { primary: "laras/deepseek-flash" },
      models: { "laras/deepseek-flash": { alias: "laras" } },
    },
  },
  models: {
    mode: "merge",
    providers: {
      laras: {
        baseUrl: "https://api.larasrouter.com/v1",
        apiKey: "${LARAS_API_KEY}",
        api: "openai-completions",
        models: [
          {
            id: "deepseek-flash",
            name: "DeepSeek Flash (Laras Router)",
            reasoning: false,
            input: ["text"],
            contextWindow: 1000000,
            maxTokens: 32000,
          },
        ],
      },
    },
  },
}

Switch to the model

Switch to it with /model laras/deepseek-flash, or run openclaw models set laras/deepseek-flash.

Good to know

The entry in models.providers.laras.models is what registers the model; the agents.defaults.models entry adds it to the model list with an alias. If you restrict models with agents.defaults.modelPolicy.allow, add laras/deepseek-flash there too.
Keep input text only: Laras Router does not accept image input. Without contextWindow, OpenClaw assumes 200,000 tokens.

Settings and screens can change between versions. If something looks different, check the OpenClaw documentation. For other tools, see all integrations.