OpenRouter

OpenRouter Provider Options

Weaver supports provider-specific configuration options for OpenRouter models. These can be set in the Provider Options (JSON) field in the chat configuration settings.

Available Provider Options

For a list of all available provider options, please refer to the OpenRouter docs. Below are some common configurations.

Reasoning

Weaver automatically enables reasoning for OpenRouter models. You can override the values by supplying a reasoning property.

{
  "reasoning": {
    // One of the following (not both):
    "effort": "high", // Can be "high", "medium", or "low" (OpenAI-style)
    "max_tokens": 2000, // Specific token limit (Anthropic-style)
    // Optional: Default is false. All models support this.
    "exclude": false // Set to true to exclude reasoning tokens from response
  }
}

Provider Routing

You can set which provider(s) to use for the model you select via the provider property

{
  "provider": {
    "order": [
      "OpenAI",
      "Together"
    ],
    "allow_fallbacks": false
  }
}

see Provider Routing for more details and options.

PDFs and Web Search

You can enable PDF parsing and OpenRouter's Exa AI web search functionality via the plugins property.

{
  "plugins": [
    {
      "id": "file-parser",
      "pdf": {
        "engine": "pdf-text", // defaults to "mistral-ocr".
      },
    },
    {
      "id": "web",
      "max_results": 5, // Defaults to 5
      "search_prompt": "Some relevant web results:"
    }
  ],
}

see PDF Support and Web Search for more details including options and defaults.