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

# LLM Calls

> A searchable, filterable table of every LLM invocation from every project, with full input and output capture.

The LLM Calls page is the single place to find any LLM invocation that has ever passed through Observatory. Use it when you know what you're looking for — a user, an agent, an error spike, a slow call — and you want to drill into the actual prompt and completion.

***

## Key features

<CardGroup cols={2}>
  <Card title="Lazy-loaded table" icon="table">
    Server-side pagination, sorting, and filtering. The page stays fast on datasets of millions of rows.
  </Card>

  <Card title="Global text search" icon="magnifying-glass">
    The search box runs across agent name, model, input, and output.
  </Card>

  <Card title="Filter by agent" icon="filter">
    Quick filter to scope the table to a single agent name.
  </Card>

  <Card title="Time range" icon="clock">
    Last Hour, 3H, 24H, 7D, 1M, 6M selector at the page header level.
  </Card>

  <Card title="Status badges" icon="circle-check">
    Success, error, in-progress, and timed-out are colour-coded at a glance.
  </Card>

  <Card title="Detail drawer" icon="sidebar">
    Click any row to open the full input, output, metadata, and cost breakdown.
  </Card>
</CardGroup>

***

## Columns

| Column         | What it shows                                                      |
| -------------- | ------------------------------------------------------------------ |
| **Status**     | `success`, `error`, `started`, or timed-out (started > 5 min).     |
| **Start time** | When the call began, in the viewer's local timezone.               |
| **Agent**      | The `@agent("name")` decorator value.                              |
| **Model**      | Provider model identifier — `gpt-4`, `claude-opus-4-7`, etc.       |
| **Duration**   | Wall-clock seconds. Below 4s is success, 4–8s warning, >8s danger. |
| **User**       | The user who triggered the call, when available in event metadata. |
| **Tokens**     | Total tokens for the call (prompt + completion).                   |
| **Requests**   | Number of HTTP requests bundled into the same logical call.        |

***

## Filtering and search

The text search box matches `contains` against the agent name, model, input, and output fields. The agent dropdown is the equivalent of an exact-match filter on the **Agent** column.

For more complex queries — status, model, tags, latency bounds, cost — use the API directly:

```http theme={"system"}
GET /api/run/?status=error&model=gpt-4&min_latency_s=8
```

See [LLMRunsQuery](../sdk/event-reporting#advanced-filtering) for the full parameter list.

***

## The detail drawer

Click any row to open the run detail drawer. It contains:

* The full prompt and completion (rendered as text or JSON depending on payload)
* The chain-of-calls tree if the run contained nested chains or tools — links to [Traces](./traces)
* Token and cost breakdown by model
* Metadata extracted by the SDK (`user_email`, custom tags, request ID)

<Tip>
  Long outputs are virtualised. Use the **Copy** button in the drawer header to grab the full payload when you need to paste it into a bug report.
</Tip>

***

## Common workflows

<AccordionGroup>
  <Accordion title="Find every failed call for one agent in the last 24 hours" icon="circle-question">
    1. Set the time range to **24H**.
    2. Pick the agent from the **Filter by agent** dropdown.
    3. Sort by **Status** to bring errors to the top.
  </Accordion>

  <Accordion title="Find calls slower than 8 seconds" icon="circle-question">
    Sort by **Duration** descending. Rows with the red `danger` badge are above the 8s threshold.
  </Accordion>

  <Accordion title="Find a call by something the user typed" icon="circle-question">
    Paste a snippet of the user's input into the search box. Matching rows surface immediately.
  </Accordion>
</AccordionGroup>

***

## Related resources

<CardGroup cols={2}>
  <Card title="Traces" icon="bars-staggered" href="./traces">
    For runs that contain nested chains and tools, see the trace view.
  </Card>

  <Card title="Analytics" icon="chart-line" href="./analytics">
    Aggregate trends across the same data.
  </Card>
</CardGroup>