The Usage section in your dashboard shows a detailed record of every request made with your API keys, along with aggregate statistics and a full transaction history.
What you can see
- Usage log — every request with model name, prompt tokens, completion tokens, charged amount, and timestamp
- Statistics — aggregate spend broken down by model
- Transactions — a ledger of all balance debits and credits with descriptions
Open Dashboard → Usage to view the usage log and statistics. Open Dashboard → Transactions for the balance ledger.
Usage log
The usage log records each individual inference request charged to your account.
curl "https://modelswitch.io/api/usage?limit=50&offset=0" \
-H "Authorization: Bearer YOUR_JWT_TOKEN"
{
"usage": [
{
"id": 1,
"model": "gpt-4o",
"promptTokens": 523,
"completionTokens": 214,
"chargedAmount": 0.00345,
"createdAt": "2024-01-15T14:22:05Z"
}
]
}
Aggregate statistics
Get totals and a per-model breakdown of your spending:
curl https://modelswitch.io/api/usage/stats \
-H "Authorization: Bearer YOUR_JWT_TOKEN"
The response includes an overall stats object and a byModel array with per-model token counts and costs.
Transaction history
Transactions represent movements in your account balance — charges for API usage and credits from top-ups.
curl "https://modelswitch.io/api/transactions?limit=50&offset=0" \
-H "Authorization: Bearer YOUR_JWT_TOKEN"
{
"transactions": [
{
"id": 1,
"amount": -0.00345,
"type": "usage",
"description": "gpt-4o inference",
"createdAt": "2024-01-15T14:22:05Z"
}
]
}
Negative amounts are charges. Positive amounts are balance top-ups.
Both /api/usage and /api/transactions support pagination via limit and offset query parameters. The maximum value for limit is 50.
# Second page of results
curl "https://modelswitch.io/api/usage?limit=50&offset=50" \
-H "Authorization: Bearer YOUR_JWT_TOKEN"
Create a separate API key for each project. Since the usage log records which key was used for each request, you can filter by key to track spending per project.