---
title: "Create agent runs"
description: "Run job:agent.runs.create through the looot API: 1 provider, from $0.15 per call. Inputs, prices, output shape and code for curl, JavaScript and Python."
sidebar:
  hidden: true
---

{/* Generated by scripts/generate-api-reference.mjs from data/api-reference.json. Do not edit. */}

Create agent runs.

Job id `agent.runs.create`, in Web search & research (Research agents). As of 2026-10-01, 1 provider serve it through 1 tool. Run `job:agent.runs.create` and looot picks one of them; with `fallback` on, a miss moves on to the next. See [Jobs](/concepts/jobs).

## Inputs

This job has one tool, so it takes that tool's fields. Any key you send passes through unchanged.

| Name | Type | Required | Description |
| --- | --- | --- | --- |
| `budget` | string | no | Optional per-run spending limit for the metered `auto` and `max` efforts. Runs that finish early may cost less than the limit. |
| `dataSources` | array | no | Exa Connect data providers to enable for the run. Each entry enables all of that provider's tools. |
| `effort` | string | no | Cost and reasoning effort preference for the run. `auto` lets Exa choose the appropriate effort. `max` is the highest-effort public beta tier for work where completeness and thoroughness matter more than latency or cost, including large list building, deep multi-source research, and criteria that are hard to verify. |
| `input` | string | no | Records to process and records or entities to exclude from the answer. |
| `metadata` | string | no | Caller-provided metadata stored with the run. |
| `outputSchema` | string | no |   |
| `previousRunId` | string | no | Agent run ID. New run IDs are returned with the `agent_run_` prefix. |
| `query` | string | yes | Natural-language question or instructions for the request. |
| `systemPrompt` | string | no | Additional instructions that guide generated output or agent behavior. Use this for source preferences, novelty constraints, duplication constraints, or other behavior guidance. Example: "Prefer official sources and avoid duplicate results." |

## Providers and prices

| Provider | Tool | Price |
| --- | --- | --- |
| [Exa](/providers/exa) | [`exa-agent-runs-post`](/reference/tools/exa-agent-runs-post) | $0.15 per call |

A call that fails at the provider costs $0. See [What's free](/money/free-and-failures).

## Output

The run's `result` holds the provider's answer. Signed in, `looot inspect exa-agent-runs-post` prints its fields.

## Run it

Every call needs your API token in `LOOOT_TOKEN`; [Sign in](/get-started/sign-in#use-the-token-in-scripts-and-agents) shows how to get one. Each run also needs a new idempotency key, so a retry never pays twice. With `wait: 30` the answer comes back inline when the run ends within 30 seconds. Otherwise you get the running run back: poll `GET /v1/runs/<runId>`.

Example input with placeholder values:

<CodeGroup>

```bash curl
curl -X POST "https://api.looot.ai/v1/runs" \
  -H "Authorization: Bearer $LOOOT_TOKEN" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: $(uuidgen)" \
  -d '{"endpointId":"job:agent.runs.create","input":{"query":"What are the most important AI infrastructure funding rounds announced this week?","metadata":{"user_id":"<user_id>","slack_thread_id":"<slack_thread_id>","slack_channel_id":"<slack_channel_id>"},"systemPrompt":"Prefer official sources and avoid duplicate results.","previousRunId":"<previousRunId>"},"wait":30}'
```

```js JavaScript
const response = await fetch("https://api.looot.ai/v1/runs", {
  method: "POST",
  headers: {
    Authorization: `Bearer ${process.env.LOOOT_TOKEN}`,
    "Content-Type": "application/json",
    "Idempotency-Key": crypto.randomUUID(),
  },
  body: JSON.stringify({
    endpointId: "job:agent.runs.create",
    input: {
      query: "What are the most important AI infrastructure funding rounds announced this week?",
      metadata: {
        user_id: "<user_id>",
        slack_thread_id: "<slack_thread_id>",
        slack_channel_id: "<slack_channel_id>",
      },
      systemPrompt: "Prefer official sources and avoid duplicate results.",
      previousRunId: "<previousRunId>",
    },
    wait: 30,
  }),
});
const run = await response.json();
console.log(run.status, run.result);
```

```python Python
import os
import uuid

import requests

response = requests.post(
    "https://api.looot.ai/v1/runs",
    headers={
        "Authorization": f"Bearer {os.environ['LOOOT_TOKEN']}",
        "Idempotency-Key": str(uuid.uuid4()),
    },
    json={
        "endpointId": "job:agent.runs.create",
        "input": {
            "query": "What are the most important AI infrastructure funding rounds announced this week?",
            "metadata": {
                "user_id": "<user_id>",
                "slack_thread_id": "<slack_thread_id>",
                "slack_channel_id": "<slack_channel_id>",
            },
            "systemPrompt": "Prefer official sources and avoid duplicate results.",
            "previousRunId": "<previousRunId>",
        },
        "wait": 30,
    },
    timeout=90,
)
run = response.json()
print(run["status"], run.get("result"))
```

</CodeGroup>

To pin one provider, send its tool id as `endpointId` instead of `job:agent.runs.create`.