OpenRouter

OverviewCopy Anchor Link

OpenRouter puts one OpenAI-compatible API in front of many model providers, and routes each request to one of them.

• OpenRouterLLMService: an OpenAILLMService subclass with OpenRouter's base URL and default model.

InstallationCopy Anchor Link

bun add @pipedog/openrouter

It depends on @pipedog/openai, which is installed with it.

import { OpenRouterLLMService } from "@pipedog/openrouter/llm";

PrerequisitesCopy Anchor Link

• An OpenRouter account, with credits, and an API key from OpenRouter.
• Pass the key as apiKey. It is required, and the service never falls back to OPENAI_API_KEY. The examples read it from OPENROUTER_API_KEY.

OpenRouterLLMServiceCopy Anchor Link

ConfigurationCopy Anchor Link

The other options (organization, project, defaultHeaders, serviceTier, retryOnTimeout, retryTimeoutSecs, function calling) are OpenAILLMService's. See OpenAI.

SettingsCopy Anchor Link

Runtime-updatable settings, passed as settings: new OpenRouterLLMService.Settings({...}) (OpenRouterLLMSettings). They are OpenAI's settings, with these defaults:

Every other setting keeps OpenAI's default. Unset sampling settings are left out of the request, so the provider's own defaults apply. extra carries provider-specific request fields.

UsageCopy Anchor Link

import { LLMUpdateSettingsFrame } from "@pipedog/core/frames/frames";
import { OpenRouterLLMService } from "@pipedog/openrouter/llm";

const llm = new OpenRouterLLMService({
  apiKey: process.env.OPENROUTER_API_KEY ?? "",
  settings: new OpenRouterLLMService.Settings({
    model: "openai/gpt-4.1",
    systemInstruction: "You are a helpful assistant in a voice conversation.",
    temperature: 0.7,
  }),
});

Constrain which providers serve the request with OpenRouterLLMService.ProviderPreferences. Its options are camelCase (allowFallbacks, preferredMaxLatency, ...); options it doesn't know are sent as given:

const routed = new OpenRouterLLMService({
  apiKey: process.env.OPENROUTER_API_KEY ?? "",
  settings: new OpenRouterLLMService.Settings({
    model: "anthropic/claude-sonnet-4",
    provider: new OpenRouterLLMService.ProviderPreferences({
      sort: "latency",
      preferredMaxLatency: { p90: 1.5 },
      allowFallbacks: true,
    }),
  }),
});

Change settings mid-call with an LLMUpdateSettingsFrame. The next completion uses them:

// Queue it on the worker, or push it from a processor upstream of the service.
const update = new LLMUpdateSettingsFrame({
  delta: new OpenRouterLLMService.Settings({ temperature: 0.2, maxTokens: 300 }),
});

NotesCopy Anchor Link

• Runs a completion for each LLMContextFrame, between an LLMFullResponseStartFrame and an LLMFullResponseEndFrame, and streams the reply as LLMTextFrames, as OpenAILLMService does. Function calls, token usage metrics and runInference work the same way.
• "developer" messages are sent as "user" messages, since the API doesn't support that role (supportsDeveloperRole is false).
• For Gemini models, every system message after the first is sent as a user message.
• A provider in extra.extra_body wins over the provider setting.

Runtime notesCopy Anchor Link

Generated from pipecat.runtimes and pipecat.compat in packages/openrouter/package.json: ✅ means the compat smoke test passes on that runtime.

• The client is the official openai JS SDK (through @pipedog/openai), which talks through fetch. It works on Bun, Node, Deno and Cloudflare Workers.
• The deprecated model constructor option is not ported. Use settings.
• Tested against a local OpenAI-compatible mock server, not against OpenRouter itself.