Skip to main content
GET
/
api
/
v1
/
costs
Cost Analytics
curl --request GET \
  --url https://api.example.com/api/v1/costs
{
  "summary": {
    "weekly_cost": {
      "value": 123,
      "trend_pct": 123
    }
  },
  "daily": [
    {
      "date": "<string>",
      "cost": 123
    }
  ],
  "hourly": [
    {
      "hour": "<string>",
      "cost": 123
    }
  ],
  "by_model": [
    {
      "model": "<string>",
      "tokens": 123,
      "share_pct": 123,
      "estimated_cost": 123
    }
  ]
}

​
Overview

The costs endpoint provides detailed cost analysis across time and models. It returns summary statistics, hourly/daily cost timeseries, and a breakdown of costs by model. Costs are calculated using model pricing data and token counts. Negative costs (invalid pricing) are sanitized to zero.

​
Authentication

Requires session authentication (cookie) or API key via X-API-Key header.

​
Query Parameters

​
range
string
default:"7d"
Time range for analytics. Valid values: 1h, 6h, 24h, 7d, 30dNote: Default is 7d for cost analytics (differs from other endpoints).
​
agent_name
string
Filter analytics to a specific agent. If omitted, returns data for all agents.

​
Response Schema

​
summary
object
Cost summary for the selected time range
​
daily
array
Daily cost timeseries
​
hourly
array
Hourly cost timeseries
​
by_model
array
Cost and token breakdown by model, ordered by token usage (highest first)

​
Example Request

cURL
curl -X GET "https://api.manifest.build/api/v1/costs?range=30d" \
  -H "X-API-Key: your-api-key"

​
Example Response

{
  "summary": {
    "weekly_cost": {
      "value": 12.48,
      "trend_pct": -8.5
    }
  },
  "daily": [
    {
      "date": "2026-02-03",
      "cost": 0.38
    },
    {
      "date": "2026-02-04",
      "cost": 0.42
    },
    {
      "date": "2026-02-05",
      "cost": 0.51
    },
    {
      "date": "2026-02-06",
      "cost": 0.39
    },
    {
      "date": "2026-02-07",
      "cost": 0.44
    },
    {
      "date": "2026-02-08",
      "cost": 0.36
    },
    {
      "date": "2026-02-09",
      "cost": 0.48
    }
  ],
  "hourly": [
    {
      "hour": "2026-02-09 00:00:00",
      "cost": 0.018
    },
    {
      "hour": "2026-02-09 01:00:00",
      "cost": 0.022
    },
    {
      "hour": "2026-02-09 02:00:00",
      "cost": 0.015
    }
  ],
  "by_model": [
    {
      "model": "claude-4.5-sonnet",
      "tokens": 3250000,
      "share_pct": 62.3,
      "estimated_cost": 8.12
    },
    {
      "model": "gpt-4o",
      "tokens": 1450000,
      "share_pct": 27.8,
      "estimated_cost": 3.18
    },
    {
      "model": "gemini-2.0-flash",
      "tokens": 520000,
      "share_pct": 9.9,
      "estimated_cost": 1.18
    }
  ]
}

​
Notes

  • Cost calculation: Costs are computed as (input_tokens * input_price + output_tokens * output_price) using model pricing from the model_pricing table.
  • Pricing data: Model prices are synced from Anthropic’s API and manually updated for other providers. Missing pricing results in null costs.
  • Cost sanitization: Negative costs (data errors) are treated as 0 and excluded from aggregations.
  • Both timeseries returned: Similar to /tokens, this endpoint returns both hourly and daily arrays in every response.
  • Caching: Responses are cached for 5 minutes (see DASHBOARD_CACHE_TTL_MS).
  • Multi-tenancy: Data is automatically filtered by authenticated user.
  • Default range: Unlike other analytics endpoints, costs default to 7d (weekly view).

Build docs developers (and LLMs) love

Get started for freeTalk to us