Key status

GET /v1/key — limits, current-window usage and what's left, without opening the dashboard.

GEThttps://iranrouter.com/v1/key

This key's limits, current-window usage and what is left — without opening the dashboard. Enough for an agent to check headroom before starting heavy work.

JSON

{
  "object": "key",
  "plan": "pro",
  "limits": {
    "requests_per_minute": 60,
    "requests_per_hour": 1000,
    "requests_per_day": 10000,
    "tokens_per_minute": 200000,
    "monthly_budget_rial": null
  },
  "usage": {
    "requests_this_hour": 42,
    "requests_today": 380,
    "spend_mtd_rial": 1840000
  },
  "remaining": {
    "requests_this_hour": 958,
    "requests_today": 9620,
    "budget_rial": null
  }
}

null means unlimited, not zero. An agent that reads null as zero stops itself for no reason.

How the limits work →

GEThttps://iranrouter.com/v1/generation?id={request_id}

One past request of this account by its x-request-id: tokens by class, counts (images, searches, OCR pages), Rial cost, latency and status — the numbers the dashboard shows. The row lands a few seconds after the answer; an immediate lookup may be a 404.

Shell

curl "https://iranrouter.com/v1/generation?id=3f6c...-request-id" -H "Authorization: Bearer ir-..."

JSON

{
  "id": "3f6c...", "object": "generation", "model": "openai/gpt-4o-mini", "status": "success", "streamed": true,
  "usage": { "prompt_tokens": 24, "completion_tokens": 108, "total_tokens": 132,
             "prompt_tokens_details": { "cached_tokens": 0, "cache_write_tokens": 0, "audio_tokens": 0, "image_tokens": 0 },
             "completion_tokens_details": { "reasoning_tokens": 0, "audio_tokens": 0, "image_tokens": 0 },
             "cost_rial": 1840, "discount_pct": 0, "estimated": false },
  "latency": { "ttft_ms": 412, "total_ms": 2210 }
}