OpenRouter Wrapper

A thin pass-through Actor that proxies OpenRouter chat completions through Apify's apify/openrouter Standby Actor.

apify/openrouter only accepts calls that originate from inside the Apify platform (caller type APIFY_ACTOR). This Actor exists to satisfy that requirement: it runs as a normal Apify Actor, and internally makes the call to openrouter.apify.actor on your behalf, so it works from any external caller using the standard run-sync-get-dataset-items API.

Input

{
  "prompt": "Your prompt text",
  "model": "openai/gpt-4o-mini",
  "reasoning_effort": "low",
  "response_format": "json_object"
}

• prompt (required) — the user message to send.
• model — OpenRouter model slug, e.g. openai/gpt-4o-mini, anthropic/claude-3.5-haiku. Defaults to openai/gpt-4o-mini.
• reasoning_effort — optional: none | low | medium | high. Omitted or none disables reasoning.
• response_format — optional: set to json_object to force JSON-mode output.

Output

A single dataset item:

{ "output": "the model's response text" }

On failure, the item instead has output: null and an error field with the upstream error body.

Authentication

Uses your Apify API token automatically (APIFY_TOKEN is injected into every Actor run by the platform) — no separate OpenRouter account or key needed. Billing flows through your Apify account via apify/openrouter's pay-per-event pricing.