Serverless Usage API

Query your own Serverless token usage and estimated cost.

Use GET /v1/usage/me to query usage attributed to the user that owns the API key. The endpoint returns daily buckets, per-model token totals, request counts, and estimated cost.

Run

curl -s "https://api.wafer.ai/v1/usage/me?period=7d&endpoint=pass.wafer.ai" \
  -H "Authorization: Bearer <SERVERLESS_API_KEY>"

Parameters

• period (optional): 1d, 7d, or 30d
• start_at and end_at (optional): explicit ISO timestamps instead of period
• endpoint (optional): restrict usage to one endpoint, such as pass.wafer.ai
• pass_account_id (optional): restrict usage to one pass account UUID

Use either period or start_at plus end_at. If no endpoint is provided, usage is aggregated across endpoints visible to the key’s owning user.

Response Shape

{
  "user_id": "00000000-0000-0000-0000-000000000000",
  "pass_account_id": null,
  "period_start": "2026-05-15T20:00:00+00:00",
  "period_end": "2026-05-22T20:00:00+00:00",
  "days": [
    {
      "date": "2026-05-22T00:00:00+00:00",
      "input_tokens": 120000,
      "output_tokens": 24000
    }
  ],
  "models": [
    {
      "model": "GLM-5.1",
      "input_tokens": 120000,
      "output_tokens": 24000,
      "cache_read_tokens": 80000,
      "request_count": 42,
      "estimated_cost_cents": 123
    }
  ],
  "total_input_tokens": 120000,
  "total_output_tokens": 24000,
  "total_cache_read_tokens": 80000,
  "total_requests": 42,
  "total_estimated_cost_cents": 123
}

Key Fields

• days: token usage buckets over the requested period
• models: per-model token totals, cache-read tokens, request counts, and estimated cost
• total_input_tokens, total_output_tokens, total_cache_read_tokens: token totals across all returned models
• total_requests: total requests attributed to the user in the window
• total_estimated_cost_cents: estimated cost in cents for the returned usage

Errors

• 401: missing or invalid API key
• 400: invalid pass_account_id or invalid time window
• 429: rate limit exceeded