> ## Documentation Index
> Fetch the complete documentation index at: https://docs.gatelit.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# Request logs

> Browse and filter the audit trail of every request routed through your Gatelit gateway.

The **Request Logs** page shows a live audit trail of every request your gateway has processed. Logs are ordered with the most recent entries first and are paginated at 50 rows per page.

## What each log entry shows

| Field               | Description                                                                  |
| ------------------- | ---------------------------------------------------------------------------- |
| **Time**            | Timestamp of when the request was received                                   |
| **Status**          | Outcome of the request: `success` or `error`                                 |
| **Subject**         | The identity associated with the request (user or service key)               |
| **Model**           | Provider and model name (for example, `openai / gpt-4o`)                     |
| **In / Out tokens** | Input token count and output token count, separated by a slash               |
| **Cost**            | Estimated request cost in USD, calculated from the model's per-token pricing |
| **Duration**        | Total round-trip time in milliseconds                                        |
| **Fallback**        | Indicates whether the request was served by a fallback provider key          |

Additional fields are stored per log entry and available when debugging:

* **Request ID** — a unique identifier for the request; see [finding a specific request](#finding-a-specific-request) below
* **Auth mode** — how the request was authenticated: `GatelitSigned`, `Bearer`, or `GatelitKey`
* **Streaming** — whether the response was streamed
* **Time to first token** — latency from request start to the first streamed token, in milliseconds (streaming requests only)
* **Prompt ID** — the ID of the saved prompt template used, if the request was linked to one

## Filtering logs

Use the filter controls above the table to narrow down entries:

* **Status** — show only requests with a specific outcome (success or error)
* **Provider** — show only requests routed to a specific provider (OpenAI, Anthropic, Google, or Mistral)
* **Subject** — type a partial subject string to search for requests from a specific user or service key

Filters combine — you can filter by both status and provider at the same time. Click **Clear** to reset all active filters.

## Finding a specific request

Every gateway response includes an `x-gatelit-request-id` header. If a client reports an issue with a specific call, copy the value of that header and paste it into the **Subject** filter to locate the matching log entry quickly.

```http theme={null}
HTTP/2 200
x-gatelit-request-id: 01J4KZXP8TMFNVEQBSRYD6WCH3
content-type: application/json
```

<Note>
  Log retention is determined by your organization's plan. Older entries may no longer appear in the dashboard after the retention window has passed. Contact support if you need access to logs beyond the visible range.
</Note>